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 .
Preliminary Amendments
Claims 1-7 have been amended. Claims 1-7 are currently pending and have been considered by the Examiner.
Specification
The abstract of the disclosure is objected to because it exceeds 150 words in length. A corrected abstract of the disclosure is required and must be presented on a separate sheet, apart from any other text. See MPEP § 608.01(b).
The title of the invention is not descriptive. A new title is required that is clearly indicative of the invention to which the claims are directed.
Claim Objections
Claims 2-3 and 6 are objected to because of the following informalities:
In claim 2, line 2 Examiner suggests replacing “a bidirectional LSTM” with “a bidirectional long short-term memory (LSTM)”.
In claim 3, line 2, Examiner suggests replacing “an LSTM” with “a long short-term memory (LSTM)”.
In claim 6, line 3, Examiner suggests adding a colon after the term “execute”. Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-7 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
In claim 1, the limitation in lines 3-4 renders the claim indefinite for the following reasons. It is unclear if “a series data set set” means a set of series data sets. The string of symbols “X = {Xd} d∈D” renders the claim indefinite because it is unclear what this string means. The string “d∈D” is not defined, and it is unclear if this means “d is an element of D”.
Specification paragraphs [0011]-[0012] appear to define Xd (subscript d) as a time-series data set of a task d. It is unclear if “Xd” as recited in the claim means Xd. Examiner treats the limitation in lines 3-4 to mean a time series data set for a task d which is an element of D.
Claim 1 recites the limitation "the task d" in line 5. There is insufficient antecedent basis for this limitation in the claim. It is unclear if “a task d∈D” in line 4 provides sufficient antecedent basis for “the task d”. Examiner treats the limitation in line 5 as having antecedent basis from “a task d∈D” (a task d which is an element of D) in line 4.
Claims 2-5 and 7 are rejected for failing to cure the deficiencies of claim 1 upon which they depend.
In claim 2, the entire limitation in lines 3-4 renders the claim indefinite because it is unclear if this limitation is supposed to mean generating a structure of the bidirectional LSTM including its latent layers, or generating an output from each latent layer of the bidirectional LSTM. Examiner treats the limitation to mean generating, at each time, an output from each latent layer of the bidirectional LSTM as the task vector.
In claim 3, the limitation “generating each latent layer of the LSTM” in line 4 renders the claim indefinite because it is unclear if this limitation is supposed to mean generating a structure of the LSTM including its latent layers, or generating an output from each latent layer of the LSTM. Examiner treats the limitation to mean generating an output from each latent layer of the LSTM.
Claim 6 is rendered indefinite because it recites the same indefinite limitations as method claim 1.
Claim 7 recites the limitation "the learning method" in lines 3-4. There is insufficient antecedent basis for this limitation in the claim. Examiner treats this limitation as “a learning method”.
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 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1-5 recite a method, claim 6 recites an apparatus comprising a processor (an apparatus), and claim 7 recites a non-transitory computer-readable recording medium (a product). Each of a method, an apparatus, and a product falls under one of the four statutory categories of patent eligible subject matter.
Claim 1
Step 2A Prong 1: Sampling the task d from the task set D and then sampling a first subset from a series data set Xd corresponding to the task d and a second subset from a set obtained by excluding the first subset from the series data set Xd is an observation mental process which can reasonably be performed in the human mind with the aid of pencil and paper.
Generating a task vector representing characteristics of the first subset is a mathematical calculation. In specification paragraphs [0022]-[0023], Equation 1 is a formula for calculating a task vector is hnt.
Calculating, from the task vector and series data included in the second subset, a predicted value of each value included in the series data is a mathematical calculation. Specification paragraphs [0027]-[0028] disclose calculating a predicted value x̂t+1 in Equations 3 and 4.
Updating learning target parameters including the parameters of the first neural network and the parameters of the second neural network using an error between each value included in the series data and the predicted value corresponding to each value is a mathematical calculation. In specification paragraphs [0032]-[0034], Equations 5 and 6 are formulas for updating learning target parameters. The claim recites an abstract idea.
Step 2A Prong 2: A computer including a memory and processor for executing a learning method amounts to generic computer components for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
Receiving a series data set set X = {Xd} d∈D composed of series data sets Xd for learning in a task d∈D when a task set is set as D amounts to mere data-gathering, an insignificant pre-solution activity under MPEP 2106.05(g).
Using parameters of a first neural network amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
Using parameters of a second neural network amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, alone or in combination, do not integrate the abstract ideas into a practical application as they are mere insignificant extra solution activities as disclosed in combination with generic computer functions that are implemented to perform the abstract ideas disclosed above. The claim is directed to an abstract idea.
Step 2B: A computer including a memory and processor for executing a learning method amounts to generic computer components for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
Receiving a series data set set X = {Xd} d∈D composed of series data sets Xd for learning in a task d∈D when a task set is set as D is analogous to receiving data over a network, which the courts have recognized as a well-understood, routine, conventional activity under MPEP 2106.05(d)(II).
Using parameters of a first neural network amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
Using parameters of a second neural network amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
The additional elements as disclosed above, in combination with the abstract ideas, are not sufficient to amount to significantly more than the abstract ideas as they are well-understood, routine and conventional activities as disclosed in combination with generic computer functions that are implemented to perform the abstract ideas disclosed above. The claim is not patent eligible.
Claim 2 incorporates the rejection of claim 1.
Step 2A Prong 1: The abstract ideas of claim 1 are incorporated.
Step 2A Prong 2 and Step 2B: The first neural network is a bidirectional LSTM amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f).
Generating includes generating each latent layer at each time of the bidirectional LSTM as the task vector amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f). The claim is not patent eligible.
Claim 3 incorporates the rejection of claim 1.
Step 2A Prong 1: The abstract ideas of claim 1 are incorporated. Generating each latent layer of the LSTM at each time as a vector representing characteristics of the series data included in the second subset is a mathematical calculation. Specification paragraphs [0025]-[0026] disclose calculating a query vector zt in Equation 2.
Calculating the predicted value of each value included in the series data from the task vector and the vector representing the characteristics of the series data is a mathematical calculation. Specification paragraphs [0027]-[0028] disclose calculating the predicted value x̂t+1 based on task vector hnt and query vector zt using Equations 3 and 4.
Step 2A Prong 2 and Step 2B: The second neural network includes an LSTM amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f). The claim is not patent eligible.
Claim 4 incorporates the rejection of claim 3.
Step 2A Prong 1: The abstract ideas of claim 3 are incorporated. Calculating the predicted value of each value included in the series data through the neural network having the attention mechanism is a mathematical calculation. Specification paragraphs [0027]-[0028] disclose calculating a predicted value x̂t+1 based on zt in Equations 3 and 4.
Step 2A Prong 2 and Step 2B: The second neural network includes a neural network having an attention mechanism, and using this neural network amounts to mere instructions for applying the abstract ideas on a generic computer under MPEP 2106.05(f). The claim is not patent eligible.
Claim 5 incorporates the rejection of claim 1.
Step 2A Prong 1: The abstract ideas of claim 1 are incorporated. Calculating the error using an expected test error or a negative log likelihood is a mathematical calculation, and updating the learning target parameters using the calculated error is a mathematical calculation. In specification paragraphs [0032]-[0034], Equations 5 and 6 are formulas for updating learning target parameters.
Step 2A Prong 2 and Step 2B: The claim does not recite any additional elements which, alone or in combination, would integrate the abstract ideas into a practical application or which, in combination with the abstract ideas, would be sufficient to amount to significantly more than the abstract ideas. The claim is not patent eligible.
Claim 6 recites an apparatus which implements the same features as the method of claim 1 and is therefore rejected for at least the same reasons.
Claim 7 recites an non-transitory computer-readable recording medium which implements the same features as the method of claim 1 and is therefore rejected for at least the same reasons.
In Step 2A Prong 2 and Step 2B, a non-transitory computer-readable recording medium having computer-readable instructions stored thereon, which when executed, cause a computer including a memory and a processor to execute the learning method amount to generic computer components for applying the abstract ideas on a generic computer under MPEP 2106.05(f). The claims is not patent eligible.
Claim Rejections - 35 USC § 103
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.
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 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.
Claims 1-3 and 5-7 are rejected under 35 U.S.C. 103 as being unpatentable over Banerjee et al. (US 20200352461 A1) in view of Sehovac et al. (“Forecasting Building Energy Consumption with Deep Learning: A Sequence to Sequence Approach”).
Regarding claim 1, Banerjee teaches: A learning method, executed by a computer including a memory and processor, the method comprising: ([0027], lines 5-11)
receiving a series data set set X = {Xd} d∈D composed of series data sets Xd for learning in a task d∈D when a task set is set as D; ([0031], [0033], lines 1-10, and [0034], lines 12-17 discloses acquiring an ECG that includes an R-R interval time series and a P wave time series. A series data set set X = {Xd} is the ECG, and the number of tasks is one.)
sampling the task d from the task set D and then sampling a first subset from a series data set Xd corresponding to the task d and a second subset from a set obtained by excluding the first subset from the series data set Xd; ([0033], lines 1-10, and [0034], lines 12-17, where a first subset is an R-R interval time series, and a second subset is a P wave time series.)
generating a task vector representing characteristics of the first subset using parameters of a first neural network; ([0032]-[0033]; [0035], line 16 to “intervals” in line 20 discloses a bidirectional LSTM network which receives R-R intervals time series. A task vector includes a hidden vector sequence ht, and parameters include all weights W and biases b of the bidirectional LSTM network.)
calculating, from the task vector and series data included in the second subset, a predicted value of each value included in the series data using parameters of a second neural network; and ([0035], lines 16-end, [0036], [0038]-[0039]. A predicted value is either “AF” or “Non-AF”. This prediction characterizes each time series data points in the P waves time series. A “second neural network” is LSTM network 110. The calculating is based on the merged output states of the two LSTM networks 108 and 110, and thus calculating the predicted value uses parameters of the second neural network, LSTM 110.)
updating learning target parameters including the parameters of the first neural network and the parameters of the second neural network using an error between [a correct label]
However, Banerjee does not explicitly teach: an error between each value included in the series data and the predicted value corresponding to each value.
But Sehovac teaches: an error between each value included in the series data and the predicted value corresponding to each value. (Page 111, col. 2, subsection B, lines 1-20 teaches usage series data. Page 112, col. 2, final 5 lines and Page 113, col. 1, lines 1-5 teaches an error is a difference between an actual value and a predicted value. The limitation of “each value included in the series data” is each actual target value to be predicted, and “the predicted value corresponding to each value” is a predicted value.)
It would have been obvious to a person having ordinary skill in the art before the effective filing date to have applied Sehovac’s error metric to a prediction by Banerjee’s network architecture disclosed by Fig. 4. A motivation for the combination is to train the LSTM to forecast a time series of data.
Regarding claim 2, the combination of Banjeree and Sehovac teaches: The learning method according to claim 1,
Banjeree teaches: wherein the first neural network is a bidirectional LSTM, and the generating includes generating each latent layer at each time of the bidirectional LSTM as the task vector. ([0032]-[0033]; [0035], line 16 to “intervals” in line 20 discloses a bidirectional LSTM network which receives R-R intervals time series. Latent layer outputs include a hidden vector sequence ht.)
Regarding claim 3, the combination of Banjeree and Sehovac teaches: The learning method according to claim 1,
Banjeree teaches: wherein the second neural network includes an LSTM, and the calculating includes generating each latent layer of the LSTM at each time as a vector representing characteristics of the series data included in the second subset, and ([0032], [0034]; [0035], from “LSTM” in line 20 to line 24 discloses an LSTM network 110 which receives P waves time series. Latent layer outputs include a hidden vector sequence ht.)
calculating the predicted value of each value included in the series data from the task vector and the vector representing the characteristics of the series data. ([0036], [0038]-[0039] teaches calculating a predicted value of “AF” or “Non-AF” based on the hidden vector sequences from the bidirectional LSTM and the LSTM.)
Regarding claim 5, the combination of Banjeree and Sehovac teaches: The learning method according to claim 1,
Banjeree teaches: wherein the updating includes calculating the error using an expected test error or a negative log likelihood, and updating the learning target parameters using the calculated error. ([0040], lines 16-end, [0042], lines 2-4 below the “var” formula, and [0043] discloses calculating cross entropy loss during training and minimizing the loss by updating the model parameters. The entire dataset was randomly partitioned into training, validation, and testing. The training loss is “an expected test error” because it is the same error one would expect if test inputs were input into the model before training finishes.)
Claim 6 recites a learning apparatus which implements the same features as the learning method of claim 1 and is therefore rejected for at least the same reasons.
Banjeree teaches: A learning apparatus comprising: a memory; and a processor configured to execute ([0027], lines 5-11)
Claim 7 is taught by the combination of Banjeree and Sehovac. Banjeree teaches: A non-transitory computer-readable recording medium having computer-readable instructions stored thereon, which when executed, cause a computer including a memory and a processor to execute the learning method according to claim 1. ([0009], lines 1-6)
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Banerjee et al. (US 20200352461 A1) in view of Sehovac et al. (“Forecasting Building Energy Consumption with Deep Learning: A Sequence to Sequence Approach”) and Sehovac et al. (“Deep Learning for Load Forecasting: Sequence to Sequence Recurrent Neural Networks With Attention”), hereinafter Sehovac II.
Regarding claim 4, the combination of Banjeree and Sehovac teaches: The learning method according to claim 3,
However, Banjeree and Sehovac do not explicitly teach: wherein the second neural network includes a neural network having an attention mechanism, and the calculating includes calculating the predicted value of each value included in the series data through the neural network having the attention mechanism.
But Sehovac II teaches: wherein the second neural network includes a neural network having an attention mechanism, and (Page 3, col. 1, section C, line 1 to col. 2, equation (10); and Page 5, col. 2, section C, lines 1-2)
the calculating includes calculating the predicted value of each value included in the series data through the neural network having the attention mechanism. (Page 9, col. 1, equation (26) and lines 1-2 below equation (27).)
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated Sehovac II’s attention mechanism into the neural network of Banerjee and Sehovac. A motivation for the combination is that the attention mechanism alleviates the burden of connecting encoder and decoder. (Sehovac II’s Abstract, lines 10-11)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Li et al. (US 20180157638 A1) at [0077], equation 9 teaches a concatenated hidden output for a bi-directional LSTM.
Dick et al. (US 20200164517 A1) at [0064] teaches training an LSTM with a dense fully-connected layer in an end-to-end manner.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Asher H. Jablon whose telephone number is (571)270-7648. The examiner can normally be reached Monday - Friday, 9:00 am - 6:00 pm.
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, Abdullah Al Kawsar can be reached at (571)270-3169. 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.
/A.H.J./Examiner, Art Unit 2127
/ABDULLAH AL KAWSAR/Supervisory Patent Examiner, Art Unit 2127