DETAILED ACTION
This Office Action is sent in response to the Applicant’s Communication received on 12/29/2022 for application number 18/003,786. The Office hereby acknowledges receipt of the following and placed of record in file: Specification, Drawings, Abstract, Oath/Declaration, IDS, and Claims.
Preliminary amendments to the claims and specification were received.
Claims 1-4 and 6 are pending.
Claims 1-4 and 6 are amended.
Claim 5 is canceled.
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 .
Claim Objections
Claims 1, 3, and 6 are objected to because of the following informalities:
All recitations of “an artificial intelligence” and “the artificial intelligence” should followed by “model,” “system,” “network,” or another relevant term
“the output” should read “an output”
“the activation functions” should read “activation functions”
“the neurons” should read “neurons”
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.
Claim 4 is 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.
The term “at least some” in claim 4 renders the claim indefinite. The term “at least some” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention.
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 1-4 and 6 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Claims 1 and 2 are directed towards a method. Claims 3 and 4 are directed towards a device. Claim 6 is directed towards a non-transitory computer-readable medium. Therefore, all claims are directed towards one of the four categories of patent eligible subject matter.
Claim 3 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim does not fall within at least one of the four categories of patent eligible subject matter because the claimed device does not demonstrate any structural recitations. Claim 3 recites a “device” claim comprising only software components such as “an artificial intelligence,” “a monitor for monitoring,” and “a monitor model,” and do not comprise any physical or tangible structure. MPEP 2106.03 states: Non-limiting examples of claims that are not directed to any of the statutory categories include: Products that do not have a physical or tangible form, such as information (often referred to as "data per se") or a computer program per se (often referred to as "software per se") when claimed as a product without any structural recitations.
Claim 1
Step 2A Prong 1:
Claim 1 recites:
“classifying input data [using the working model];” Classifying input data is an action that can be performed mentally with the aid of pen and paper, and is therefore a mental process.
Step 2A Prong Two
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“An automated method for generating a monitor model of a monitor having a first neural network for monitoring a working model of an artificial intelligence having a second neural network;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“using the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“training the monitoring model using a representative volume of data of the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“wherein the representative volume of data is used as the input for the first neural network and wherein the representative volume of data is formed by the output of the activation functions of the neurons of the second neural network;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“An automated method for generating a monitor model of a monitor having a first neural network for monitoring a working model of an artificial intelligence having a second neural network;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“using the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“training the monitoring model using a representative volume of data of the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“wherein the representative volume of data is used as the input for the first neural network and wherein the representative volume of data is formed by the output of the activation functions of the neurons of the second neural network;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
Even when considered in combination, these additional elements represent mere instructions to apply an exception and therefore do not provide an inventive concept. The claim is ineligible.
Claim 2
Step 2A Prong Two
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“wherein training the monitor model includes: putting the training data into the working model which is fully trained with training data;” Mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity (MPEP 2106.05(g)).
“running said training data through the working model once more for analysis;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“wherein the monitor model learns a set of valid activation patterns and/or activation paths which the working model can achieve;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“wherein training the monitor model includes: putting the training data into the working model which is fully trained with training data;” Mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g). The additional element of “receiving” does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of receiving steps amounts to no more than mere data gathering. This element amounts to receiving data over a network and are well-understood, routine, conventional activity. See MPEP 2106.05(d), subsection II (i). This cannot provide an inventive concept.
“running said training data through the working model once more for analysis;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“wherein the monitor model learns a set of valid activation patterns and/or activation paths which the working model can achieve;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
Even when considered in combination, these additional elements represent mere instructions to apply an exception and therefore do not provide an inventive concept. The claim is ineligible.
Claim 3
Step 2A Prong 1:
Claim 3 recites:
“wherein the artificial intelligence is configured to perform a classification of input data [using the working model];” Performing a classification of input is an action that can be performed mentally with the aid of pen and paper, and is therefore a mental process.
Step 2A Prong Two
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“A device comprising: artificial intelligence; and a monitor for monitoring a working model of the artificial intelligence;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“using the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“wherein a monitor model of the monitor is trained with a representative volume of data of the working model;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
“wherein the monitor has an artificial first neural network;” “the artificial intelligence has an artificial second neural network;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“the input for the first neural network is formed by the representative volume of data;” “the representative volume of data is formed by the output of the activation functions of the neurons of the second neural network;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“A device comprising: artificial intelligence; and a monitor for monitoring a working model of the artificial intelligence;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“using the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“wherein a monitor model of the monitor is trained with a representative volume of data of the working model;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
“wherein the monitor has an artificial first neural network;” “the artificial intelligence has an artificial second neural network;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“the input for the first neural network is formed by the representative volume of data;” “the representative volume of data is formed by the output of the activation functions of the neurons of the second neural network;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
Even when considered in combination, these additional elements represent mere instructions to apply an exception and therefore do not provide an inventive concept. The claim is ineligible.
Claim 4
Step 2A Prong Two
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“the monitor model is trained by putting at least some of the training data into the working model which is fully trained with training data and running said training data through the working model once more for analysis;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
“the monitor model learns a set of valid activation patterns and/or activation paths which the working model can achieve;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“the monitor model is trained by putting at least some of the training data into the working model which is fully trained with training data and running said training data through the working model once more for analysis;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
“the monitor model learns a set of valid activation patterns and/or activation paths which the working model can achieve;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
Even when considered in combination, these additional elements represent mere instructions to apply an exception and therefore do not provide an inventive concept. The claim is ineligible.
Claim 6
Step 2A Prong 1:
Claim 6 recites:
“classifying input data [using the working model];” Classifying input data is an action that can be performed mentally with the aid of pen and paper, and is therefore a mental process.
Step 2A Prong Two
This judicial exception is not integrated into a practical application because the additional elements are as follows:
“A non-transitory computer-readable medium storing a computer program wherein the computer program is loadable into a storage device of a device to perform an automated method for generating a monitor model of a monitor having a first neural network for monitoring a working model of an artificial intelligence having a second neural network;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“using the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“training the monitoring model using a representative volume of data of the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)).
“wherein the representative volume of data is used as the input for the first neural network and wherein the representative volume of data is formed by the output of the activation functions of the neurons of the second neural network;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)).
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are as follows:
“A non-transitory computer-readable medium storing a computer program wherein the computer program is loadable into a storage device of a device to perform an automated method for generating a monitor model of a monitor having a first neural network for monitoring a working model of an artificial intelligence having a second neural network;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“using the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“training the monitoring model using a representative volume of data of the working model;” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (MPEP 2106.05(f)) which cannot provide an inventive concept.
“wherein the representative volume of data is used as the input for the first neural network and wherein the representative volume of data is formed by the output of the activation functions of the neurons of the second neural network;” The limitation amounts to merely indicating a field of use or technological environment in which to apply a judicial exception. This does not amount to significantly more than the exception itself (MPEP 2106.05(h)) and cannot provide an inventive concept.
Even when considered in combination, these additional elements represent mere instructions to apply an exception and therefore do not provide an inventive concept. The claim is ineligible.
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 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) 1-3 and 6 are rejected under 35 U.S.C. 103 as being unpatentable over キャノイ、 et al. (JP2017509952A, see attached translation), hereinafter JPA, and Cheng et al. (Runtime Monitoring Neuron Activation Patterns, published 2019), hereinafter Cheng.
Regarding claim 1, JPA teaches,
An automated method for generating a monitor model of a monitor having a first neural network for monitoring a working model of an artificial intelligence having a second neural network [Para 0004, The method includes monitoring a first neural network with a second neural network; Para 0091, Neurons in the shadow network may monitor neurons in the target network; Para 0088, shadow neural networks may be created manually or automatically in hardware and/or software], the method comprising:
Classifying input data using the working model [Para 0019, Every neuron in level 106 can generate an output spike 110 based on the corresponding synthesized input signal. The output spikes 110 may be transferred to neurons at another level using another network of synaptic connections; Para 0020, Biological synapses can be classified as either electrical or chemical synapses]; and
Training the monitoring model using a representative volume of data of the working model [Para 0095, when a condition is detected within the neurons and/or synapses of the target network, the shadow network is reconfigured based on the detected condition. Specifically, the shadow network may reconfigure its size, parameters, and/or thresholds based on the detected conditions. The reconfiguration may be automated or may be performed by a user of the shadow network. Additionally, the shadow network may be trainable based on user input and/or monitoring may be performed by the shadow network];
JPA teaches the above limitations of claim 1 including the first neural network and the second neural network (Para 0004 and 0091).
JPA does not teach wherein representative volume of data is used as the input for first network and wherein the representative volume of data is formed by the output of the activation functions of the neurons of second neural network.
Cheng teaches,
wherein representative volume of data is used as the input for first network [Fig 1(a), encircled portion; Sect I, para 2, one records the neuron activation patterns for close-to-output neural network layers for all correctly predicted data used in the training process
PNG
media_image1.png
563
771
media_image1.png
Greyscale
]; and
wherein the representative volume of data is formed by the output of the activation functions of the neurons of second neural network [Sect II, para 5, Definition 1 (Neuron activation pattern): Given a neural network with input in and the l-th layer being ReLU, pat(f(l)(in)), the neuron activation pattern at layer l, is defined as follows: pat(f(l)(in)) := (prelu(v1), . . . , prelu(vdl )) where (v1, . . . , vdl) = f(l)(in) is the output from layer l, and prelu : R → {0, 1} captures the activation cases].
Cheng is analogous to the claimed invention as they both relate to monitoring neural networks using monitoring models. Therefore, it would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified JPA’s teachings to incorporate the teachings of Cheng and provide data formed by activation functions in the neural network and inputting the data into the monitoring model in order for [Cheng, Sect I, para 1] the network to extrapolate from what it learns (or remembers) from the training data, as similar data has not appeared in the training process, and improve the machine learning system.
Regarding claim 2, JPA-Cheng teach the limitations of claim 1 including training the monitor model (JPA, Para 0095).
Cheng further teaches,
putting the training data into the working model which is fully trained with training data; and running said training data through the working model once more for analysis; wherein the monitor model learns a set of valid activation patterns and/or activations paths which the working model can achieve [Abstract, after the standard training process, one creates a monitor by feeding the training data to the network again in order to store the neuron activation patterns in abstract form].
Cheng is analogous to the claimed invention as they both relate to monitoring neural networks using monitoring models. Therefore, it would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified JPA’s teachings to incorporate the teachings of Cheng and provide the monitor model learning activation patterns of the working model’s training input in order to [Cheng, Abstract] report a significant portion of misclassifications to be not supported by training with a small false-positive rate.
Regarding claim 3, JPA teaches,
A device comprising: an artificial intelligence; and a monitor for monitoring a working model of artificial intelligence [Para 0004, The method includes monitoring a first neural network with a second neural network; Para 0091, Neurons in the shadow network may monitor neurons in the target network; Para 0088, shadow neural networks may be created manually or automatically in hardware and/or software];
Wherein the artificial intelligence is configured to perform a classification of input data using the working model [Para 0019, Every neuron in level 106 can generate an output spike 110 based on the corresponding synthesized input signal. The output spikes 110 may be transferred to neurons at another level using another network of synaptic connections; Para 0020, Biological synapses can be classified as either electrical or chemical synapses]; and
Wherein a monitor model of the monitor is trained with a representative volume of data of the working model [Para 0095, when a condition is detected within the neurons and/or synapses of the target network, the shadow network is reconfigured based on the detected condition. Specifically, the shadow network may reconfigure its size, parameters, and/or thresholds based on the detected conditions. The reconfiguration may be automated or may be performed by a user of the shadow network. Additionally, the shadow network may be trainable based on user input and/or monitoring may be performed by the shadow network];
wherein the monitor has an artificial first neural network [Para 0004, The method includes monitoring a first neural network with a second neural network];
the artificial intelligence has an artificial second neural network [Para 0004, The method includes monitoring a first neural network with a second neural network];
JPA teaches the above limitations of claim 1 including the first neural network, the second neural network (Para 0004 and 0091), and the representative volume of data (Para 0024).
JPA does not teach the input for first network is formed by representative volume of data and the representative volume of data is formed by the output of the activation functions of the neurons of second neural network.
Cheng teaches,
The input for first network is formed by representative volume of data [Fig 1(a), encircled portion; Sect I, para 2, one records the neuron activation patterns for close-to-output neural network layers for all correctly predicted data used in the training process
PNG
media_image1.png
563
771
media_image1.png
Greyscale
]; and
the representative volume of data is formed by the output of the activation functions of the neurons of second neural network [Sect II, para 5, Definition 1 (Neuron activation pattern): Given a neural network with input in and the l-th layer being ReLU, pat(f(l)(in)), the neuron activation pattern at layer l, is defined as follows: pat(f(l)(in)) := (prelu(v1), . . . , prelu(vdl )) where (v1, . . . , vdl) = f(l)(in) is the output from layer l, and prelu : R → {0, 1} captures the activation cases].
Cheng is analogous to the claimed invention as they both relate to monitoring neural networks using monitoring models. Therefore, it would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified JPA’s teachings to incorporate the teachings of Cheng and provide data formed by activation functions in the neural network and inputting the data into the monitoring model in order for [Sect I, para 1] the network to extrapolate from what it learns (or remembers) from the training data, as similar data has not appeared in the training process, and improve the machine learning system.
Regarding claim 6, JPA teaches,
A non-transitory computer-readable medium storing a computer program wherein the computer program is loadable into a storage device of a device to perform an automated method for generating a monitor model of a monitor having a first neural network for monitoring a working model of an artificial intelligence having a second neural network [Para 0004, The method includes monitoring a first neural network with a second neural network; Para 0006, a computer program product for generating events having a non-transitory computer-readable medium is disclosed. The computer-readable medium has non-transitory program code recorded thereon that, when executed by a processor, causes the processor to perform an operation of monitoring a first neural network with a second neural network. The program code also causes the processor to generate an event based on the monitoring; Para 0091, Neurons in the shadow network may monitor neurons in the target network; Para 0088, shadow neural networks may be created manually or automatically in hardware and/or software; Para 0124, A storage medium may be coupled to the processor such that the processor can read information from, and write information to, the storage medium], the method comprising:
Classifying input data using the working model [Para 0019, Every neuron in level 106 can generate an output spike 110 based on the corresponding synthesized input signal. The output spikes 110 may be transferred to neurons at another level using another network of synaptic connections; Para 0020, Biological synapses can be classified as either electrical or chemical synapses]; and
Training the monitoring model using a representative volume of data of the working model [Para 0095, when a condition is detected within the neurons and/or synapses of the target network, the shadow network is reconfigured based on the detected condition. Specifically, the shadow network may reconfigure its size, parameters, and/or thresholds based on the detected conditions. The reconfiguration may be automated or may be performed by a user of the shadow network. Additionally, the shadow network may be trainable based on user input and/or monitoring may be performed by the shadow network];
JPA teaches the above limitations of claim 1 including the first neural network and the second neural network (Para 0004 and 0091).
JPA does not teach wherein representative volume of data is used as the input for first network and wherein the representative volume of data is formed by the output of the activation functions of the neurons of second neural network.
Cheng teaches,
wherein representative volume of data is used as the input for first network [Fig 1(a), encircled portion; Sect I, para 2, one records the neuron activation patterns for close-to-output neural network layers for all correctly predicted data used in the training process
PNG
media_image1.png
563
771
media_image1.png
Greyscale
]; and
wherein the representative volume of data is formed by the output of the activation functions of the neurons of second neural network [Sect II, para 5, Definition 1 (Neuron activation pattern): Given a neural network with input in and the l-th layer being ReLU, pat(f(l)(in)), the neuron activation pattern at layer l, is defined as follows: pat(f(l)(in)) := (prelu(v1), . . . , prelu(vdl )) where (v1, . . . , vdl) = f(l)(in) is the output from layer l, and prelu : R → {0, 1} captures the activation cases].
Cheng is analogous to the claimed invention as they both relate to monitoring neural networks using monitoring models. Therefore, it would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified JPA’s teachings to incorporate the teachings of Cheng and provide data formed by activation functions in the neural network and inputting the data into the monitoring model in order for [Sect I, para 1] the network to extrapolate from what it learns (or remembers) from the training data, as similar data has not appeared in the training process, and improve the machine learning system.
Claim(s) 4 is rejected under 35 U.S.C. 103 as being unpatentable over Cheng in view of JPA, and in further view of Yuan et al. (US 11599813 B1), hereinafter Yuan.
Regarding claim 4, JPA-Cheng teach the limitations of claim 3 including the monitor model is trained (JPA, Para 0095).
Cheng further teaches,
working model which is fully trained with training data; Monitor model learns a set of valid activation pattern and/or activation paths which the working model can achieve [Abstract, after the standard training process, one creates a monitor by feeding the training data to the network again in order to store the neuron activation patterns in abstract form].
Cheng is analogous to the claimed invention as they both relate to monitoring neural networks using monitoring models. Therefore, it would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified JPA’s teachings to incorporate the teachings of Cheng and provide learning a set of valid activation pattern and/or activation paths in order to [Cheng, Abstract] report a significant portion of misclassifications to be not supported by training with a small false-positive rate.
JPA-Cheng do not teach monitor model putting at least some of the training data into the working model and running said training data through the working model once more for analysis.
Yuan teaches,
Monitor model putting at least some of the training data into the working model and running said training data through the working model once more for analysis [Fig 3, emboxed portion; Col 12, lines 43-67 and Col 13, lines 1-7, Model training and evaluation 350 may include training models by launching multiple training jobs with different algorithms, tuning parameters, and evaluating the best performing model for promotion. The one or more model training and evaluation 350 tasks may provide one or more model artifacts 355 to the one or more model deployment and inference tasks 360… The one or more model deployment and inference tasks 360 may generate inference results 365 that are usable by one or more model monitoring tasks 370… The one or more model monitoring tasks 370 may provide inference metrics 375 to the one or more model training and evaluation 350 tasks].
PNG
media_image2.png
624
690
media_image2.png
Greyscale
Yuan is analogous to the claimed invention as they both relate to efficient model monitoring and training. Therefore, it would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Cheng and JPA’s teachings to incorporate the teachings of Yuan and provide the monitoring model putting training data back into the working model for further analysis [Yuan, Col 13, lines 2-5] in order to prevent model degradation.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SYED RAYHAN AHMED whose telephone number is (571)270-0286. The examiner can normally be reached Mon-Fri ET.
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, David Yi can be reached at (571) 270-7519. 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.
/SYED RAYHAN AHMED/Examiner, Art Unit 2126
/DAVID YI/Supervisory Patent Examiner, Art Unit 2126