DETAILED ACTION
This action is responsive to the Application/amendment filed on 07/06/2026. Claims 1-7, 9-16 and 18-20 are pending in the case. Claims 1, 9, and 19 are independent claims. Claims 1-2, 5, 9-16 and 18-20 are amended while claims 8 and 17 have been cancelled.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Claim Objections
Claims 1, 10 and 19 are objected to because of the following informalities: each of the listed claims have a typo referring to a time series as a timer series. Appropriate correction is required.
Claim Rejections - 35 USC § 112
Claims 1-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, because the specification, while being enabling for a SIREN model receiving a time series with at least one value that is missing, does not reasonably provide enablement for generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity… wherein the irregularity includes at least one value that is missing. The specification does not enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the invention commensurate in scope with these claims. While the specification cites multiple times the SIREN model can receive the time series and other information with a missing value ([0010], “…receiving, by the at least one processor, first information that relates to a latent vector representation of a first time series for which at least one value is missing” and other disclosures such as [0023], “receive first information that relates to a latent vector representation of a first time series for which at least one value is missing”, Figure 4 showing receiving a vector representation of missing values) Examiner could not find any disclosure of a SIREN generating a time series where the time series has values missing. While claims 1, 10 and 19 cite the limitation, dependent claims inherent the deficiencies of parent claims and therefore claims 1-20 are rejected.
Claims 1-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, because the specification, while being enabling for a using latent values as variables during training and backpropagated, does not reasonably provide enablement for updating, by the at least one processor, the SIREN model for auto-decoding, wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model. The specification does not enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the invention commensurate in scope with these claims. While the specification cites using latent values as variables during training and backpropagated ([0079] "...In this setup, the latent values are treated as variables during training (and so backpropagated), and then optimized again during inference for a given time series, when all other network variables are fixed.") Examiner could not find any disclosure using the imputed value as a variable to be used in training of the SIREN . While claims 1, 10 and 19 cite the limitation, dependent claims inherent the deficiencies of parent claims and therefore claims 1-20 are rejected.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-7, 9-16 and 18-20 are rejected under 35 U.S.C. 101 because claims are directed towards abstract ideas/mental processes without significantly more.
Regarding Claim 1:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites generating…a first time series including an irregularity, wherein the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, and wherein the irregularity includes at least one value that is missing which, under the broadest reasonable interpretation, covers performance of the limitation in the mind with the aid of pen and paper. The limitations encompass choosing using evaluation to judge/organize observations according to time. See 2106.04.(a)(2).III.C
The claim recites modulating…based on the first information…a set of sine activation amplitudes of the functional representation of the first time series which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass choosing different values for a variable. See 2106.04.(a)(2).III.C
The claim recites imputing…based on the set of network weights and a result of the modulating…the at least one missing value of the first time series which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating a pattern and using judgement to determine a value. See 2106.04.(a)(2).III.C.
The claim recites generating…based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites mapping the generated set of network weights to the SIREN model to instantiate the first time series from which imputation is carried out which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating a pattern and using judgement to determine a value. See 2106.04.(a)(2).III.C.
The claim recites updating… the SIREN model… which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating using judgment to change parameters of a model. See 2106.04.(a)(2).III.C.
Subject Matter Eligibility Analysis Step 2A Prong 2:
the method being implemented by at least one processor and a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN) which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
training, by the at least one processor, a SIREN model with historical data which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
receiving…first information that relates to a latent vector representation of the first time series for which at least one value is missing which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
generating, by a neural network model in the hypernet and based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by the hypernet which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by a neural network model in the hypernet which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by the at least one processor which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
for auto-decoding wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model which specifies a particular technological environment in which the abstract idea is to take place, i.e. a field of use (see MPEP 2106.05(h)
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) (b) (d) (e) (f) and (g) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
Additional element (c) of obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Additional elements (h) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation merely specifies a field of use in which the abstract idea is to take place, i.e. a field of use (see MPEP 2106.05(h)).
The additional element(s) (a) (b) (c) (d) (e) (f) (g) and (h) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 2:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim does not contain elements that would warrant a Step 2A Prong 1 analysis.
Subject Matter Eligibility Analysis Step 2A Prong 2:
the modulating of the set of sine activation amplitudes is performed by using another neural network model which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) (b) and (c) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
The additional element(s) (a) (b) and (c) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 3:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites assessing an accuracy of the imputing which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass using judgement to determine an opinion on the similarity of two values/sets of values. See 2106.04.(a)(2).III.C.
The claim recites obtaining…a first metric that relates to a mean- squared error between the imputed at least one missing value of the first time series and at least one ground truth value that corresponds to the at least one missing value of the first time series which are abstract ideas(Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites obtaining…a second metric that relates to a maximum error between the imputed atleast one missing value of the first time series and an evaluation of a model output that corresponds to the at least one missing value of the first time series which are abstract ideas(Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites obtaining…a third metric that relates to a Euclidean distance in a feature space between the imputed at least one missing value of the first time series and the at least one ground truth value that corresponds to the at least one missing value of the first time series which are abstract ideas(Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
The claim does not contain elements that would warrant a Step 2A Prong 2 analysis.
Subject Matter Eligibility Analysis Step 2B:
The claim does not include any additional element, when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 4:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites wherein the modulating is further based on the second information which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations appears to clarify using information for the mental process of modulating a variable and encompass choosing different values for a variable based on a data. See 2106.04.(a)(2).III.C
Subject Matter Eligibility Analysis Step 2A Prong 2:
receiving second information that relates to frequency modes of a superset of data that includes the first time series and at least a second time series which merely recites insignificant extra-solution activity of data gathering (see MPEP 2106.05(g)).
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 ).
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 5:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim recites with respect to a predetermined loss function which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
Subject Matter Eligibility Analysis Step 2A Prong 2:
training the sinusoidal representation network model which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
using historical data and optimizing the sinusoidal representation network model which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) and (b) is merely clarifying the type of information that is obtained as a network input. Obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362)
The additional element(s) (a) and (b) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 6:
The rejection of claim 1 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim does not contain elements that would warrant a Step 2A Prong 1 analysis.
Subject Matter Eligibility Analysis Step 2A Prong 2:
wherein the first time series comprises a univariate time series which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) is merely clarifying the type of information that is obtained as a network input. Obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 7:
The rejection of claim 6 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim does not contain elements that would warrant a Step 2A Prong 1 analysis.
Subject Matter Eligibility Analysis Step 2A Prong 2:
wherein the univariate time series comprises a time series that relates to stock market data which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) is merely clarifying the type of information that is obtained as a network input. Obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362)
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 9:
The rejection of claim 8 is incorporated and further claim recites further additional
elements/limitations:
Subject Matter Eligibility Analysis Step 2A Prong 1:
The claim does not contain elements that would warrant a Step 2A Prong 1 analysis.
Subject Matter Eligibility Analysis Step 2A Prong 2:
wherein the multivariate time series comprises one from among a time series that relates to yield rate curve data, a time series that relates to weather forecasting data, and a time series that relates to medical diagnosis data which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
Additional element (a) is merely clarifying the type of information that is obtained as a network input. Obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362)
The additional element(s) (a) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 10:
The claim recites generating…a first time series including an irregularity, wherein the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, and wherein the irregularity includes at least one value that is missing which, under the broadest reasonable interpretation, covers performance of the limitation in the mind with the aid of pen and paper. The limitations encompass choosing using evaluation to judge/organize observations according to time. See 2106.04.(a)(2).III.C
The claim recites modulating…based on the first information…a set of sine activation amplitudes of the functional representation of the first time series which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass choosing different values for a variable. See 2106.04.(a)(2).III.C
The claim recites imputing…based on the set of network weights and a result of the modulating…the at least one missing value of the first time series which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating a pattern and using judgement to determine a value. See 2106.04.(a)(2).III.C.
The claim recites generating…based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites mapping the generated set of network weights to the SIREN model to instantiate the first time series from which imputation is carried out which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating a pattern and using judgement to determine a value. See 2106.04.(a)(2).III.C.
The claim recites updating… the SIREN model… which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating using judgment to change parameters of a model. See 2106.04.(a)(2).III.C.
Subject Matter Eligibility Analysis Step 2A Prong 2:
a processor; a memory; a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN); a communication interface coupled to each of the processor and the memory, wherein the system is configured to which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
training, by the at least one processor, a SIREN model with historical data which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
receiving…first information that relates to a latent vector representation of the first time series for which at least one value is missing which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
generating, by a neural network model in the hypernet and based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by the hypernet which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by a neural network model in the hypernet which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by the at least one processor which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
for auto-decoding wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model which specifies a particular technological environment in which the abstract idea is to take place, i.e. a field of use (see MPEP 2106.05(h)
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) (b) (d) (e) (f) and (g) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
Additional element (c) of obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Additional elements (h) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation merely specifies a field of use in which the abstract idea is to take place, i.e. a field of use (see MPEP 2106.05(h)).
The additional element(s) (a) (b) (c) (d) (e) (f) (g) and (h) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 11:
The rejection of claim 10 is incorporated and further claim recites further additional elements/limitations:
Claim 11 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 2 found in claim 11.
Regarding Claim 12:
The rejection of claim 10 is incorporated and further claim recites further additional elements/limitations:
Claim 12 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 3 found in claim 12.
Regarding Claim 13:
The rejection of claim 10 is incorporated and further claim recites further additional elements/limitations:
Claim 13 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 4 found in claim 13.
Regarding Claim 14:
The rejection of claim 10 is incorporated and further claim recites further additional elements/limitations:
Claim 14 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 5 found in claim 14.
Regarding Claim 15:
The rejection of claim 10 is incorporated and further claim recites further additional elements/limitations:
Claim 15 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 6 found in claim 15.
Regarding Claim 16:
The rejection of claim 15 is incorporated and further claim recites further additional elements/limitations:
Claim 16 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 7 found in claim 16.
Regarding Claim 18:
The rejection of claim 10 is incorporated and further claim recites further additional elements/limitations:
Claim 18 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 9 found in claim 18.
Regarding Claim 19:
The claim recites generating…a first time series including an irregularity, wherein the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, and wherein the irregularity includes at least one value that is missing which, under the broadest reasonable interpretation, covers performance of the limitation in the mind with the aid of pen and paper. The limitations encompass choosing using evaluation to judge/organize observations according to time. See 2106.04.(a)(2).III.C
The claim recites modulating…based on the first information…a set of sine activation amplitudes of the functional representation of the first time series which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass choosing different values for a variable. See 2106.04.(a)(2).III.C
The claim recites imputing…based on the set of network weights and a result of the modulating…the at least one missing value of the first time series which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating a pattern and using judgement to determine a value. See 2106.04.(a)(2).III.C.
The claim recites generating…based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series which is an abstract idea (Mathematical Calculations (see MPEP 2106.04(a)(2)(I)(C))).
The claim recites mapping the generated set of network weights to the SIREN model to instantiate the first time series from which imputation is carried out which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating a pattern and using judgement to determine a value. See 2106.04.(a)(2).III.C.
The claim recites updating… the SIREN model… which, under the broadest reasonable interpretation, covers performance of the limitation in the mind. The limitations encompass evaluating using judgment to change parameters of a model. See 2106.04.(a)(2).III.C.
Subject Matter Eligibility Analysis Step 2A Prong 2:
non-transitory computer readable storage medium storing instructions for performing time series imputation, the storage medium comprising executable code which, when executed by a system including a processor modulation network, a hypernet and a sinusoidal representation network (SIREN); causes the system to which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
training, by the at least one processor, a SIREN model with historical data which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
receiving…first information that relates to a latent vector representation of the first time series for which at least one value is missing which amount to mere extra solution activity of obtaining and/or gathering data over a network, see MPEP §2106.05(g)
generating, by a neural network model in the hypernet and based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by the hypernet which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by a neural network model in the hypernet which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
by the at least one processor which merely recites a generic computer on which to perform the abstract idea, e.g. "apply it on a computer" (see MPEP 2106.05(f))
for auto-decoding wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model which specifies a particular technological environment in which the abstract idea is to take place, i.e. a field of use (see MPEP 2106.05(h)
Subject Matter Eligibility Analysis Step 2B:
Additional elements (a) (b) (d) (e) (f) and (g) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation amount to no more than mere instructions to apply the exception using a generic computer component. Please see MPEP §2106.05(f).
Additional element (c) of obtaining a network input is well understood, routine, and conventional activity of “transmitting or receiving data over a network" (see MPEP 2106.05(d)(II)(i) using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Additional elements (h) do not integrate the abstract idea into a practical application nor do the additional limitation provide significantly more than the abstract idea because the limitation merely specifies a field of use in which the abstract idea is to take place, i.e. a field of use (see MPEP 2106.05(h)).
The additional element(s) (a) (b) (c) (d) (e) (f) (g) and (h) in the claim do/does not include any additional elements , when considered separately and in combination, that amount to an integration of the judicial exception into a practical application, nor significantly more than the judicial exception for the reasons set forth in step 2A prong 2 analysis above. The claim is not patent eligible.
Regarding Claim 20:
The rejection of claim 19 is incorporated and further claim recites further additional elements/limitations:
Claim 20 is rejected under that same 101 claim analysis due to the substantially similarity of the limitations and additional elements of claim 2 found in claim 20.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-6, 10-15, and 19-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Fons et al(“HyperTime: Implicit Neural Representation for Time Series” henceforth known as Fons) in view of Mehta et al(“Modulated Periodic Activations for Generalizable Local Functional Representations” henceforth known as Mehta) and further in view of Naour al(“Time Series Continuous Modeling for Imputation and Forecasting with Implicit Neural Representations” henceforth known as Naour) with Lai et al.(“ Attribute-Associated Neuron Modeling and Missing Value Imputation for Incomplete Data” henceforth known as Lai) and Luo et al.(“ Multivariate Time Series Imputation with Generative Adversarial Networks” henceforth known as Luo)
Regarding claim 1:
Fons discloses training, by the at least one processor,(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss, and takes less than 10s in a GeForce GTX 1080 Ti GPU”) a SIREN model(Fons, Page 3, Paragraph 5, “The network is composed of fully-connected layers of dimensions 1 × 60 × 60 × 60 × 1, with sine activations (SIREN [37]”) with historical data(Fons, Page 5, Paragraph 2,“ from each dataset, train a single SIREN for each time series” where each dataset and each time series correspond to historical data)
Fons discloses receiving, by the hypernet, first information that relates to a latent vector representation of the first time series(Fons, Page 4, Figure 2 and Paragraph 2, “The Set Encoder (green network), composed of SIREN layers [37] with dimensions 2 × 128×128×40, takes as input a pair of values, corresponding to the time-coordinate t and the time series signal f(t). Each pair of input values is thus encoded into a full 40-values embedding and fed to the HyperNet decoder (red network)”)
Fons discloses generating, by a neural network model in the hypernet and based on the first information(Fons, Page 4, Paragraph 2, “The Set Encoder (green network), composed of SIREN layers [37] with dimensions 2 × 128×128×40, takes as input a pair of values, corresponding to the time-coordinate t and the time series signal f(t). Each pair of input values is thus encoded into a full 40-values embedding and fed to the HyperNet decoder (red network)” where the fully-connected layers correspond to a neural network of the HyperNet corresponds to a neural network model in the hypernet and based on the ), a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series(Fons, Page 4, Figure 2, where the HypoNet weights are generated by the neural network for the HyperNet and used in the HypoNet )
Fons discloses mapping the generated set of network weights(Fons, Page 4, Paragraph 2, “The output of the HyperNet is a one-dimensional 7500-values embedding that contains the network weights of an INR which encodes the time series data from the input.”) to the SIREN model(Fons, Page 3, Paragraph 5, “The network is composed of fully-connected layers of dimensions 1 × 60 × 60 × 60 × 1, with sine activations (SIREN [37]”) to instantiate the first time series from which imputation is carried out(Fons, Page 5, Paragraph 5, “Motivated by this, we first use a SIREN trained with the available data points to infer the missing values” where the use of available time-series points to infer missing values corresponds to imputation being carried out(See also Fons, Page 6, Paragraph 2-3, “We compare this approach with common time series imputation methods, mean, that fills missing values with the average, kNN, cubic spline and linear…Table 2 shows the comparison of SIREN with and without prior with the classical methods using different ratios of missing values”))
Fons discloses updating…the SIREN model(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss”)
Fons discloses imputing….at least one missing value of the first time series(Fons, Page 5, Paragraph 5, “Motivated by this, we first use a SIREN trained with the available data points to infer the missing values”)
Fons does not disclose the following limitations:
modulating, by the at least one processor based on the first information, a set of sine activation amplitudes of the functional representation of the first time series
and updating…the SIREN model for auto-decoding
and imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series
generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity, wherein the first time series is multivariate time series…and wherein the irregularity includes at least one value that is missing
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Mehta discloses modulating, by the at least one processor based on the first information, a set of sine activation amplitudes of the functional representation of the first time series(Mehta, Page 4, Col. 1, Paragraph 2, “the latent codes z can modulate the amplitude of the sine activations of each
References Fons and Mehta are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations using SIREN’s.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons and Mehta before him or her, to modify the SIREN of Fons to include the modulation of Mehta to adapt to different signals latent code. The suggestion/motivation for doing so would have been Mehta, Page 2, Col. 1, Paragraph 1, “functional mapping that uses two MLPs: a modulator and a synthesis network…The modulator is the key to generalization.”
Mehta discloses and updating, by the at least one processor, the SIREN model for auto-decoding(Mehta, Col. 1, Paragraph 3, “In the auto-decoder configuration, we jointly optimize the network parameters θ, and the latent codes for all the training signals…and optimizing it with the same objective used during training”)
Fons-Mehta does not disclose the following limitations:
generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity, wherein the first time series is multivariate time series…and wherein the irregularity includes at least one value that is missing
and imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Naour discloses a network architecture including a modulation network(Naour, Page 2, Paragraph 5-6, “Our proposed framework is a modulated INR that can effectively model univariate time series with multiple individuals as a continuous function of time. To the best of our knowledge, we are one of the first to adapt INR to tackle imputation and forecasting tasks simultaneously for continuous modeling”), a hypernet(Naour, Page 3, Paragraph 6, “This modification is generated by a hypernetwork, which takes the code as input and outputs the modification. Therefore, the INR and the hypernetwork are two key components of our model.”) and a sinusoidal representation network (SIREN)( Naour, Page 13, Paragraph 2, “In Table 4, we compare the imputation results obtained by the Fourier features network and the SIREN network”)
Naour discloses generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity(Naour, Page 7 Figure 3, the bottom Inference section and Paragraph 4, “Setting For a given time series x(j),
T
i
n
(
j
)
in denotes the look-back window of L points and
T
o
u
t
(
j
)
out the horizon of H points. During training, at each epoch, we randomly draw several pairs of look-back windows and horizon
T
i
n
j
∪
T
o
u
t
j
j
∈
B
” where the model output on New period
T
(
j
*
)
corresponds to a time series where the red points are the observed look-back window and not a generated value are considered an irregularity and the purple/blue are model-generated forecast points), wherein the first time series is multivariate time series (Naour, Page 9, Paragraph 2, “To extend our findings to the multi-sample multivariate case, a simple approach is to modify the output dimension of the INR network”)…and wherein the irregularity includes at least one value that is missing(Naour, Figure 3, the bottom Inference section, where the red points are the observed look-back window and not a generated value are considered an irregularity that is missing as the red points are missing from the generated portion of the time series)
Naour discloses and imputing(Naour, Page 8, Paragraph 5, “We compared TimeFlow with the following two-step processing baseline: linear interpolation handling the missing values within the partially observed look-back window,” where imputing the missing values of the time series correspond imputing…the at least one missing value of the first time(See Also Naour, Page 20, Table 13)), by the at least one processor based on the set of network weights and a result of the modulating(Naour, Page 4, Paragraphs 1-3, “In our case, the network models the time series xt as a function of a time coordinate t…To reconstruct different univariate time series…with an INR fθ, we modulate the weights θ of the coordinate-based network with a hypernetwork hw. For a time series…the hypernetwork maps a latent code…containing specific knowledge about the time series to…the parameter space of the INR… We denote the modulated INR…where its outputs are approximations of the time series.”), the at least one missing value of the first time series(Naour, Page 3, Paragraph 6, “we use a recent DL approach called implicit neural representation (INR), which can learn from discrete data a neural network fθ(t) that can output the value at any time t”)
References Fons-Mehta and Naour are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta and Naour before him or her, to modify the inference of Fons-Mehta to include the imputation of previous missing data of Naour to better handle real-world data as real-world data has missing values. The suggestion/motivation for doing so would have been Naour, Page 2, Paragraph 1, “Practical considerations, such as the installation of new sensors, or legal restrictions, such as the EU GDPR, may prevent the simultaneous availability of all samples over time.”
Fons-Mehta-Naour does not disclose the following limitations:
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Lai discloses wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model(Lai, Page 4, Col. 2, Paragraph 1, “training scheme initializes the missing value variables in incomplete records and…updates the missing value variables and the network parameters through the backpropagation algorithm. The above process is repeated for all records”)
References Fons-Mehta-Naour and Lai are analogous art because they are from the same field of endeavor if using machine learning for imputation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour and Lai before him or her, to modify the backpropagation of Fons-Mehta-Naour to include the imputation of previous missing data of Lai for more accurate prediction of missing values. The suggestion/motivation for doing so would have been Lai, Page 4, Col. 2, Paragraph 1, “The accuracy of the model will be improved with the deepening of the training, and the missing value predicted by the model will also be more accurate.”
Fons-Mehta-Naour-Lai does not disclose the following limitations:
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Luo discloses the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, (Luo, Page 2, Paragraph 4, Given a collection of multivariate time series with d dimensions, one time series X observed in T = (t0,...,tn-1), is denoted by X = (xt0 ,...,xti ,...,xtn−1 ) ∈ Rn×d, where xti is the tith observation of X…In the following example ,d=4, n=3 and “none” is missing value,
PNG
media_image1.png
55
470
media_image1.png
Greyscale
.)
References Fons-Mehta-Naour-Lai and Luo are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour-Lai and Luo before him or her, to modify time series representation of Fons-Mehta-Naour-Lai to include the matrix representation of Luo for uniform readability as it is useful for missing-data handling and is typical for handling multivariate time-series modeling.
Regarding claim 2:
The rejection of claim 1 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
Mehta further discloses the modulating of the set of sine activation amplitudes is performed by using another neural network model(Mehta, Page 3, Col. 1, Paragraph 4, “The latent codes are processed by a modulation network, which conditionally modulates the activations of a synthesis network, that acts as a template for the functional mapping”)
Regarding claim 3:
The rejection of claim 1 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
Luo discloses further comprising assessing an accuracy of the imputing by obtaining at least one from among a first metric that relates to a mean- squared error between the imputed at least one missing value of the first time series and at least one ground truth value that corresponds to the at least one missing value of the first time series(Luo, Page 6, Paragraph 1, “The imputation accuracy is defined as the mean squared error (MSE) between original values and imputed values” where the use of mean square error to show how close the reconstructed matches the original ground-truth corresponds to a first metric that relates to a mean- squared error between the imputed at least one missing value of the first time series and at least one ground truth value that corresponds to the at least one missing value of the first time series)
Due to the claim language stating at least one from among the following limitations are not quired as the first metric is found in prior art:
a second metric that relates to a maximum error between the imputed atleast one missing value of the first time series and an evaluation of a model output that corresponds to the at least one missing value of the first time series,
and a third metric that relates to a Euclidean distance in a feature space between the imputed at least one missing value of the first time series and the at least one ground truth value that corresponds to the at least one missing value of the first time series.
Regarding claim 5:
The rejection of claim 1 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
Fons further discloses optimizing the SIREN model with respect to a predetermined loss function(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss”)
Regarding claim 4:
The rejection of claim 1 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
Mehta further discloses receiving second information that relates to frequency modes of a superset of data that includes the first time series and at least a second time series, wherein the modulating is further based on the second information(Mehta, Page 4, Col. 1, Equation 4 and Paragraph 3, “As can be seen from Equations 4 and 2, the latent codes z can modulate the amplitude of the sine activations of each hidden layer in the synthesis network, through the modulation parameters αi”)
Regarding claim 6:
The rejection of claim 1 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
Naour, further discloses wherein the first time series comprises a univariate time series(Naour, Page 1, Paragraph 2, “In this paper, we focus on these univariate time series with multiple instances/samples and address two critical tasks: imputation and long-term time series forecasting (LTSF)”)
Regarding claim 10:
Fons discloses a processor; a memory…a communication interface coupled to each of the processor and the memory, wherein the processor is configured to(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss, and takes less than 10s in a GeForce GTX 1080 Ti GPU”)
Fons discloses training, by the at least one processor,(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss, and takes less than 10s in a GeForce GTX 1080 Ti GPU”) a SIREN model(Fons, Page 3, Paragraph 5, “The network is composed of fully-connected layers of dimensions 1 × 60 × 60 × 60 × 1, with sine activations (SIREN [37]”) with historical data(Fons, Page 5, Paragraph 2,“ from each dataset, train a single SIREN for each time series” where each dataset and each time series correspond to historical data)
Fons discloses receiving, by the hypernet, first information that relates to a latent vector representation of the first time series(Fons, Page 4, Figure 2 and Paragraph 2, “The Set Encoder (green network), composed of SIREN layers [37] with dimensions 2 × 128×128×40, takes as input a pair of values, corresponding to the time-coordinate t and the time series signal f(t). Each pair of input values is thus encoded into a full 40-values embedding and fed to the HyperNet decoder (red network)”)
Fons discloses generating, by a neural network model in the hypernet and based on the first information(Fons, Page 4, Paragraph 2, “The Set Encoder (green network), composed of SIREN layers [37] with dimensions 2 × 128×128×40, takes as input a pair of values, corresponding to the time-coordinate t and the time series signal f(t). Each pair of input values is thus encoded into a full 40-values embedding and fed to the HyperNet decoder (red network)” where the fully-connected layers correspond to a neural network of the HyperNet corresponds to a neural network model in the hypernet and based on the ), a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series(Fons, Page 4, Figure 2, where the HypoNet weights are generated by the neural network for the HyperNet and used in the HypoNet )
Fons discloses mapping the generated set of network weights(Fons, Page 4, Paragraph 2, “The output of the HyperNet is a one-dimensional 7500-values embedding that contains the network weights of an INR which encodes the time series data from the input.”) to the SIREN model(Fons, Page 3, Paragraph 5, “The network is composed of fully-connected layers of dimensions 1 × 60 × 60 × 60 × 1, with sine activations (SIREN [37]”) to instantiate the first time series from which imputation is carried out(Fons, Page 5, Paragraph 5, “Motivated by this, we first use a SIREN trained with the available data points to infer the missing values” where the use of available time-series points to infer missing values corresponds to imputation being carried out(See also Fons, Page 6, Paragraph 2-3, “We compare this approach with common time series imputation methods, mean, that fills missing values with the average, kNN, cubic spline and linear…Table 2 shows the comparison of SIREN with and without prior with the classical methods using different ratios of missing values”))
Fons discloses updating…the SIREN model(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss”)
Fons discloses imputing….at least one missing value of the first time series(Fons, Page 5, Paragraph 5, “Motivated by this, we first use a SIREN trained with the available data points to infer the missing values”)
Fons does not disclose the following limitations:
modulating, by the at least one processor based on the first information, a set of sine activation amplitudes of the functional representation of the first time series
and updating…the SIREN model for auto-decoding
and imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series
generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity, wherein the first time series is multivariate time series…and wherein the irregularity includes at least one value that is missing
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Mehta discloses modulating, by the at least one processor based on the first information, a set of sine activation amplitudes of the functional representation of the first time series(Mehta, Page 4, Col. 1, Paragraph 2, “the latent codes z can modulate the amplitude of the sine activations of each
References Fons and Mehta are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations using SIREN’s.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons and Mehta before him or her, to modify the SIREN of Fons to include the modulation of Mehta to adapt to different signals latent code. The suggestion/motivation for doing so would have been Mehta, Page 2, Col. 1, Paragraph 1, “functional mapping that uses two MLPs: a modulator and a synthesis network…The modulator is the key to generalization.”
Mehta discloses and updating, by the at least one processor, the SIREN model for auto-decoding(Mehta, Col. 1, Paragraph 3, “In the auto-decoder configuration, we jointly optimize the network parameters θ, and the latent codes for all the training signals…and optimizing it with the same objective used during training”)
Fons-Mehta does not disclose the following limitations:
generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity, wherein the first time series is multivariate time series…and wherein the irregularity includes at least one value that is missing
and imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Naour discloses a network architecture including a modulation network(Naour, Page 2, Paragraph 5-6, “Our proposed framework is a modulated INR that can effectively model univariate time series with multiple individuals as a continuous function of time. To the best of our knowledge, we are one of the first to adapt INR to tackle imputation and forecasting tasks simultaneously for continuous modeling”), a hypernet(Naour, Page 3, Paragraph 6, “This modification is generated by a hypernetwork, which takes the code as input and outputs the modification. Therefore, the INR and the hypernetwork are two key components of our model.”) and a sinusoidal representation network (SIREN)( Naour, Page 13, Paragraph 2, “In Table 4, we compare the imputation results obtained by the Fourier features network and the SIREN network”)
Naour discloses generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity(Naour, Page 7 Figure 3, the bottom Inference section and Paragraph 4, “Setting For a given time series x(j),
T
i
n
(
j
)
in denotes the look-back window of L points and
T
o
u
t
(
j
)
out the horizon of H points. During training, at each epoch, we randomly draw several pairs of look-back windows and horizon
T
i
n
j
∪
T
o
u
t
j
j
∈
B
” where the model output on New period
T
(
j
*
)
corresponds to a time series where the red points are the observed look-back window and not a generated value are considered an irregularity and the purple/blue are model-generated forecast points), wherein the first time series is multivariate time series (Naour, Page 9, Paragraph 2, “To extend our findings to the multi-sample multivariate case, a simple approach is to modify the output dimension of the INR network”)…and wherein the irregularity includes at least one value that is missing(Naour, Figure 3, the bottom Inference section, where the red points are the observed look-back window and not a generated value are considered an irregularity that is missing as the red points are missing from the generated portion of the time series)
Naour discloses and imputing(Naour, Page 8, Paragraph 5, “We compared TimeFlow with the following two-step processing baseline: linear interpolation handling the missing values within the partially observed look-back window,” where imputing the missing values of the time series correspond imputing…the at least one missing value of the first time(See Also Naour, Page 20, Table 13)), by the at least one processor based on the set of network weights and a result of the modulating(Naour, Page 4, Paragraphs 1-3, “In our case, the network models the time series xt as a function of a time coordinate t…To reconstruct different univariate time series…with an INR fθ, we modulate the weights θ of the coordinate-based network with a hypernetwork hw. For a time series…the hypernetwork maps a latent code…containing specific knowledge about the time series to…the parameter space of the INR… We denote the modulated INR…where its outputs are approximations of the time series.”), the at least one missing value of the first time series(Naour, Page 3, Paragraph 6, “we use a recent DL approach called implicit neural representation (INR), which can learn from discrete data a neural network fθ(t) that can output the value at any time t”)
References Fons-Mehta and Naour are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta and Naour before him or her, to modify the inference of Fons-Mehta to include the imputation of previous missing data of Naour to better handle real-world data as real-world data has missing values. The suggestion/motivation for doing so would have been Naour, Page 2, Paragraph 1, “Practical considerations, such as the installation of new sensors, or legal restrictions, such as the EU GDPR, may prevent the simultaneous availability of all samples over time.”
Fons-Mehta-Naour does not disclose the following limitations:
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Lai discloses wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model(Lai, Page 4, Col. 2, Paragraph 1, “training scheme initializes the missing value variables in incomplete records and…updates the missing value variables and the network parameters through the backpropagation algorithm. The above process is repeated for all records”)
References Fons-Mehta-Naour and Lai are analogous art because they are from the same field of endeavor if using machine learning for imputation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour and Lai before him or her, to modify the backpropagation of Fons-Mehta-Naour to include the imputation of previous missing data of Lai for more accurate prediction of missing values. The suggestion/motivation for doing so would have been Lai, Page 4, Col. 2, Paragraph 1, “The accuracy of the model will be improved with the deepening of the training, and the missing value predicted by the model will also be more accurate.”
Fons-Mehta-Naour-Lai does not disclose the following limitations:
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Luo discloses the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, (Luo, Page 2, Paragraph 4, Given a collection of multivariate time series with d dimensions, one time series X observed in T = (t0,...,tn-1), is denoted by X = (xt0 ,...,xti ,...,xtn−1 ) ∈ Rn×d, where xti is the tith observation of X…In the following example ,d=4, n=3 and “none” is missing value,
PNG
media_image1.png
55
470
media_image1.png
Greyscale
.)
References Fons-Mehta-Naour-Lai and Luo are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour-Lai and Luo before him or her, to modify time series representation of Fons-Mehta-Naour-Lai to include the matrix representation of Luo for uniform readability as it is useful for missing-data handling and is typical for handling multivariate time-series modeling.
Regarding claim 11:
The rejection of claim 10 incorporated in claim 11. Claim 11 is rejected under the same rationale as set forth in the rejection of claim 2.
Regarding claim 12:
The rejection of claim 10 incorporated in claim 12. Claim 12 is rejected under the same rationale as set forth in the rejection of claim 3.
Regarding claim 13:
The rejection of claim 10 incorporated in claim 13. Claim 13 is rejected under the same rationale as set forth in the rejection of claim 4.
Regarding claim 14:
The rejection of claim 10 incorporated in claim 14. Claim 14 is rejected under the same rationale as set forth in the rejection of claim 5.
Regarding claim 15:
The rejection of claim 10 incorporated in claim 15. Claim 15 is rejected under the same rationale as set forth in the rejection of claim 6.
Regarding claim 19:
Fons discloses a non-transitory computer readable storage medium storing instructions for performing time series imputation, the storage medium comprising executable code which, when executed by a system including a processor(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss, and takes less than 10s in a GeForce GTX 1080 Ti GPU”)
Fons discloses training, by the at least one processor,(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss, and takes less than 10s in a GeForce GTX 1080 Ti GPU”) a SIREN model(Fons, Page 3, Paragraph 5, “The network is composed of fully-connected layers of dimensions 1 × 60 × 60 × 60 × 1, with sine activations (SIREN [37]”) with historical data(Fons, Page 5, Paragraph 2,“ from each dataset, train a single SIREN for each time series” where each dataset and each time series correspond to historical data)
Fons discloses receiving, by the hypernet, first information that relates to a latent vector representation of the first time series(Fons, Page 4, Figure 2 and Paragraph 2, “The Set Encoder (green network), composed of SIREN layers [37] with dimensions 2 × 128×128×40, takes as input a pair of values, corresponding to the time-coordinate t and the time series signal f(t). Each pair of input values is thus encoded into a full 40-values embedding and fed to the HyperNet decoder (red network)”)
Fons discloses generating, by a neural network model in the hypernet and based on the first information(Fons, Page 4, Paragraph 2, “The Set Encoder (green network), composed of SIREN layers [37] with dimensions 2 × 128×128×40, takes as input a pair of values, corresponding to the time-coordinate t and the time series signal f(t). Each pair of input values is thus encoded into a full 40-values embedding and fed to the HyperNet decoder (red network)” where the fully-connected layers correspond to a neural network of the HyperNet corresponds to a neural network model in the hypernet and based on the ), a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series(Fons, Page 4, Figure 2, where the HypoNet weights are generated by the neural network for the HyperNet and used in the HypoNet )
Fons discloses mapping the generated set of network weights(Fons, Page 4, Paragraph 2, “The output of the HyperNet is a one-dimensional 7500-values embedding that contains the network weights of an INR which encodes the time series data from the input.”) to the SIREN model(Fons, Page 3, Paragraph 5, “The network is composed of fully-connected layers of dimensions 1 × 60 × 60 × 60 × 1, with sine activations (SIREN [37]”) to instantiate the first time series from which imputation is carried out(Fons, Page 5, Paragraph 5, “Motivated by this, we first use a SIREN trained with the available data points to infer the missing values” where the use of available time-series points to infer missing values corresponds to imputation being carried out(See also Fons, Page 6, Paragraph 2-3, “We compare this approach with common time series imputation methods, mean, that fills missing values with the average, kNN, cubic spline and linear…Table 2 shows the comparison of SIREN with and without prior with the classical methods using different ratios of missing values”))
Fons discloses updating…the SIREN model(Fons, Page 3, Paragraph 6, “Training of the network is done in a supervised manner, with MSE loss”)
Fons discloses imputing….at least one missing value of the first time series(Fons, Page 5, Paragraph 5, “Motivated by this, we first use a SIREN trained with the available data points to infer the missing values”)
Fons does not disclose the following limitations:
modulating, by the at least one processor based on the first information, a set of sine activation amplitudes of the functional representation of the first time series
and updating…the SIREN model for auto-decoding
and imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series
generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity, wherein the first time series is multivariate time series…and wherein the irregularity includes at least one value that is missing
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Mehta discloses modulating, by the at least one processor based on the first information, a set of sine activation amplitudes of the functional representation of the first time series(Mehta, Page 4, Col. 1, Paragraph 2, “the latent codes z can modulate the amplitude of the sine activations of each
References Fons and Mehta are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations using SIREN’s.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons and Mehta before him or her, to modify the SIREN of Fons to include the modulation of Mehta to adapt to different signals latent code. The suggestion/motivation for doing so would have been Mehta, Page 2, Col. 1, Paragraph 1, “functional mapping that uses two MLPs: a modulator and a synthesis network…The modulator is the key to generalization.”
Mehta discloses and updating, by the at least one processor, the SIREN model for auto-decoding(Mehta, Col. 1, Paragraph 3, “In the auto-decoder configuration, we jointly optimize the network parameters θ, and the latent codes for all the training signals…and optimizing it with the same objective used during training”)
Fons-Mehta does not disclose the following limitations:
generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity, wherein the first time series is multivariate time series…and wherein the irregularity includes at least one value that is missing
and imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Naour discloses a system including a processor and a network architecture including a modulation network (Naour, Page 2, Paragraph 5-6, “Our proposed framework is a modulated INR that can effectively model univariate time series with multiple individuals as a continuous function of time. To the best of our knowledge, we are one of the first to adapt INR to tackle imputation and forecasting tasks simultaneously for continuous modeling”), a hypernet(Naour, Page 3, Paragraph 6, “This modification is generated by a hypernetwork, which takes the code as input and outputs the modification. Therefore, the INR and the hypernetwork are two key components of our model.”) and a sinusoidal representation network (SIREN)( Naour, Page 13, Paragraph 2, “In Table 4, we compare the imputation results obtained by the Fourier features network and the SIREN network”)
Naour discloses generating, by executing the trained SIREN model via the at least one processor, a first time series including an irregularity(Naour, Page 7 Figure 3, the bottom Inference section and Paragraph 4, “Setting For a given time series x(j),
T
i
n
(
j
)
in denotes the look-back window of L points and
T
o
u
t
(
j
)
out the horizon of H points. During training, at each epoch, we randomly draw several pairs of look-back windows and horizon
T
i
n
j
∪
T
o
u
t
j
j
∈
B
” where the model output on New period
T
(
j
*
)
corresponds to a time series where the red points are the observed look-back window and not a generated value are considered an irregularity and the purple/blue are model-generated forecast points), wherein the first time series is multivariate time series (Naour, Page 9, Paragraph 2, “To extend our findings to the multi-sample multivariate case, a simple approach is to modify the output dimension of the INR network”)…and wherein the irregularity includes at least one value that is missing(Naour, Figure 3, the bottom Inference section, where the red points are the observed look-back window and not a generated value are considered an irregularity that is missing as the red points are missing from the generated portion of the time series)
Naour discloses and imputing(Naour, Page 8, Paragraph 5, “We compared TimeFlow with the following two-step processing baseline: linear interpolation handling the missing values within the partially observed look-back window,” where imputing the missing values of the time series correspond imputing…the at least one missing value of the first time(See Also Naour, Page 20, Table 13)), by the at least one processor based on the set of network weights and a result of the modulating(Naour, Page 4, Paragraphs 1-3, “In our case, the network models the time series xt as a function of a time coordinate t…To reconstruct different univariate time series…with an INR fθ, we modulate the weights θ of the coordinate-based network with a hypernetwork hw. For a time series…the hypernetwork maps a latent code…containing specific knowledge about the time series to…the parameter space of the INR… We denote the modulated INR…where its outputs are approximations of the time series.”), the at least one missing value of the first time series(Naour, Page 3, Paragraph 6, “we use a recent DL approach called implicit neural representation (INR), which can learn from discrete data a neural network fθ(t) that can output the value at any time t”)
References Fons-Mehta and Naour are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta and Naour before him or her, to modify the inference of Fons-Mehta to include the imputation of previous missing data of Naour to better handle real-world data as real-world data has missing values. The suggestion/motivation for doing so would have been Naour, Page 2, Paragraph 1, “Practical considerations, such as the installation of new sensors, or legal restrictions, such as the EU GDPR, may prevent the simultaneous availability of all samples over time.”
Fons-Mehta-Naour does not disclose the following limitations:
wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model
a network architecture including a modulation network, a hypernet and a sinusoidal representation network (SIREN)
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Lai discloses wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model(Lai, Page 4, Col. 2, Paragraph 1, “training scheme initializes the missing value variables in incomplete records and…updates the missing value variables and the network parameters through the backpropagation algorithm. The above process is repeated for all records”)
References Fons-Mehta-Naour and Lai are analogous art because they are from the same field of endeavor if using machine learning for imputation.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour and Lai before him or her, to modify the backpropagation of Fons-Mehta-Naour to include the imputation of previous missing data of Lai for more accurate prediction of missing values. The suggestion/motivation for doing so would have been Lai, Page 4, Col. 2, Paragraph 1, “The accuracy of the model will be improved with the deepening of the training, and the missing value predicted by the model will also be more accurate.”
Fons-Mehta-Naour-Lai does not disclose the following limitations:
the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature,
Luo discloses the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, (Luo, Page 2, Paragraph 4, Given a collection of multivariate time series with d dimensions, one time series X observed in T = (t0,...,tn-1), is denoted by X = (xt0 ,...,xti ,...,xtn−1 ) ∈ Rn×d, where xti is the tith observation of X…In the following example ,d=4, n=3 and “none” is missing value,
PNG
media_image1.png
55
470
media_image1.png
Greyscale
.)
References Fons-Mehta-Naour-Lai and Luo are analogous art because they are from the same field of endeavor if using machine learning with implicit neural representations.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour-Lai and Luo before him or her, to modify time series representation of Fons-Mehta-Naour-Lai to include the matrix representation of Luo for uniform readability as it is useful for missing-data handling and is typical for handling multivariate time-series modeling.
Regarding claim 20:
The rejection of claim 19 incorporated in claim 20. Claim 20 is rejected under the same rationale as set forth in the rejection of claim 2.
Regarding claims 7 and 16:
Claim(s) 7 and 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Fons et al(“HyperTime: Implicit Neural Representation for Time Series” henceforth known as Fons) in view of Mehta et al(“Modulated Periodic Activations for Generalizable Local Functional Representations” henceforth known as Mehta) and further in view of Naour al(“Time Series Continuous Modeling for Imputation and Forecasting with Implicit Neural Representations” henceforth known as Naour) with Lai et al.(“ Attribute-Associated Neuron Modeling and Missing Value Imputation for Incomplete Data” henceforth known as Lai), Luo et al.(“ Multivariate Time Series Imputation with Generative Adversarial Networks” henceforth known as Luo) and Mehtab et al(“Stock Price Prediction Using Convolutional Neural Networks on a Multivariate Timeseries” henceforth known Mehtab).
Regarding claim 7:
The rejection of claim 6 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
The Fons-Mehta-Naour-Lai-Luo combination does not disclose wherein the univariate time series comprises a time series that relates to stock market data, however Mehtab discloses representing stock market data as a time series(Mehtab Page 4, Col. 1, Paragraph 2, “In the first case, we design a CNN for multi-step time series forecasting using only the univariate sequence of the close_perc values” and Mehtab Page 4, Col. 1, Paragraph 3, “The multi-step time series forecasting approach is essentially an autoregression process. Whether univariate or multivariate, the prior time series data is used for forecasting the values for the next week.”)
References Fons-Mehta-Naour-Lai-Luo and Mehtab are analogous art because they are from the field of endeavor as using time series with machine learning models to forecast data.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour-Lai-Luo and Mehtab before him or her, to modify the time series of Fons-Mehta-Naour-Lai-Luo with the stock information in a time series of Mehtab as stocks naturally form a time series as they track movements over time. The suggestion/motivation for doing so would have been Mehtab Page 2, Col. 1, Paragraph 2, “We believe that past movement patterns of daily NIFTY index values can be learned by powerful machine learning and deep learning-based approaches, and that knowledge can be gainfully applied for predicting future movement NIFTY index values.”
Regarding claim 16:
The rejection of claim 15 incorporated in claim 16. Claim 16 is rejected under the same rationale as set forth in the rejection of claim 7.
Regarding claims 9 and 18:
Claim(s) 9 and 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Fons et al(“HyperTime: Implicit Neural Representation for Time Series” henceforth known as Fons) in view of Mehta et al(“Modulated Periodic Activations for Generalizable Local Functional Representations” henceforth known as Mehta) and further in view of Naour al(“Time Series Continuous Modeling for Imputation and Forecasting with Implicit Neural Representations” henceforth known as Naour) with Lai et al.(“ Attribute-Associated Neuron Modeling and Missing Value Imputation for Incomplete Data” henceforth known as Lai), Luo et al.(“ Multivariate Time Series Imputation with Generative Adversarial Networks” henceforth known as Luo) and Festag et al(“Medical multivariate time series imputation and forecasting based on a recurrent conditional Wasserstein GAN and attention” henceforth known Festag).
Regarding claim 9:
The rejection of claim 8 with prior art Fons-Mehta-Naour-Lai-Luo is incorporated and further:
The Fons-Mehta-Naour-Lai-Luo combination does not disclose wherein the multivariate time series comprises one from among a time series that relates to yield rate curve data, a time series that relates to weather forecasting data, and a time series that relates to medical diagnosis data however Festag discloses representing medical data as a time series(Festag Page 1, Abstract, “…the focus of our work lay on a…multivariate generative approach that samples infillings or forecasts…for this task, we developed…networks that consist of…encoders and decoders…and can learn the distribution of intervals from multivariate time…The presented generative probabilistic system for the imputation and forecasting of (medical) time series” and Festag Page 2, Col. 2, Paragraph 5, “The data basis for training and testing consists of multivariate time series 𝐒 ∈ R𝑛×𝑑 with 𝑑 time steps and 𝑛 channels”)
References Fons-Mehta-Naour-Lai-Luo and Festag are analogous art because they are from the field of endeavor as using time series with machine learning models to forecast and impute data.
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art, having the teachings of Fons-Mehta-Naour-Lai-Luo and Festag before him or her, to modify the time series of Fons-Mehta-Naour-Lai-Luo with the medical information in a time series of Festag as biomedical measurements naturally form a time series as they track physiological variables over time. The suggestion/motivation for doing so would have been Festag Page 1, Col. 1, Paragraph 1, “In the biomedical domain, many measured values are part of time series representing the development of the measurand over time. Prominent examples of the temporal documentation of repeated readings in the clinical context are bedside monitors that continuously output health-related data of patients.”
Regarding claim 18:
The rejection of claim 17 incorporated in claim 18. Claim 18 is rejected under the same rationale as set forth in the rejection of claim 9.
Relevant Prior Art
While not used in the rejection the following prior art was found by the Examiner:
Sitzmann et al. "Implicit neural representations with periodic activation functions" as a general SIREN core architecture prior art.
Response to Arguments
Applicant's arguments filed 07/06/2026 have been fully considered but they are not persuasive. A breakdown of arguments can be found below:
35 USC 101:
Applicant appears to argue on page 14 that the current claims are directed to a practical application and overcomes 101 and cites to several limitations (i – vi) and asserts that, since irregularity in the multivariate time series data are regularized via modulation, imputation of missing values and backpropagation, multivariate time series data may be more efficiently processed without requiring excessive usage of computing resources as conventionally required, an ordinary artisan would reasonable understand that a more efficient utilization of limited computer resources via regularization of the multivariate time series resolves the problem of computationally heavy computing resources and constrained data complexity.
The highlighted limitations are:
initially training a SIREN model with historical data;
generating, by executing the trained SIREN model, a first time series including an irregularity, wherein the first time series is multivariate time series represented by a matrix of values, wherein each column of the matrix representing an individual timer series corresponding to a feature, and wherein the irregularity includes at least one value that is missing;
modulating, based on the first information, a set of sine activation amplitudes of the functional representation of the first time series;
imputing, based on the set of network weights and a result of the modulating, the at least one missing value of the first time series;
mapping the generated set of network weights to the SIREN model to instantiate the first time series from which imputation is carried out; and
updating the SIREN model for auto-decoding, wherein the updating includes backpropagating the imputed at least one missing value as at least one variable to be included in the training of the updated SIREN model.
Applicant respectfully disagrees as the claims are silent about regularizing via modulation and data may be more efficiently processed without requiring excessive usage of computing resources as a result of the regularization. Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). Applicant appears to be interpreting a narrower claim as the current claims do not positively recite providing a technological improvement concerning using less memory or computing resources. Further, Applicant’s highlighted limitations are categorized as abstract/mental concepts and Applicant’s does not explain how any additional elements reflect the improvements or how the improvement is affected by any claimed additional elements. Applicants highlighted limitations may describe an improvement provided by the claimed abstract/mental ideas, however an improvement of an abstract/mental concept is an improvement on said abstract/mental concept and not a technologic improvement.
35 USC 103:
Applicant appears to argue on pages 16-17 that combining multiple known elements is not rendered obvious simply because each element was known independently in the prior art and there must be a “reason that would have prompted” a person of ordinary skill to make the change. Applicant argues that Sitzmann and Mehta fail to disclose or render obvious imputing, by the at least one processor based on the set of network weights and a result of the modulating, the at least one missing value of the first time series and that Naour does not cure the deficiencies of as an ordinary artisan would not reasonable equate imputation by linear interpolation of data with imputation based on a set of network weights and result of the modulating as generally recited in amended claim 1.
Applicant’s arguments have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. Specifically, Sitzmann has been removed as prior art and Fonz has been added for the following limitation:
receiving, by the hypernet, first information that relates to a latent vector representation of the first time series
generating, by a neural network model in the hypernet and based on the first information, a set of network weights that is usable by the SIREN model for obtaining a functional representation of the first time series
mapping the generated set of network weights to the SIREN model to instantiate the first time series from which imputation is carried out
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHARLES JEFFREY JONES JR whose telephone number is (703)756-1414. The examiner can normally be reached Monday - Friday 8:00 - 5:00 EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kakali Chaki can be reached at 571-272-3719. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/C.J.J./Examiner, Art Unit 2122
/MICHAEL H HOANG/PRIMARY EXAMINER, Art Unit 2122
