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 .
Claims 1-20 are presented for examination.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier.
Claim limitations "an encoder-decoder model comprising an encoder and a decoder”;
“generator is configured to receive random sample vectors and generate synthetic representations”, and “decoder is configured to decode the synthetic representations" in claim 12, have been interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because they use a generic placeholder
“generator" coupled with functional language "is configured to receive random sample vectors and generate synthetic representations”, and
“decoder" coupled with functional language "is configured to decode the synthetic representations" in claim 12.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
A review of the specification shows that the following appears to be the corresponding structure described in the specification for the 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph limitation: according to [0032, 0107-0109], "generator" and “decoder” are implemented as hardware/ processor(s) executing software modules to perform the instructions of the modules
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
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.
Claim(s) 1-20 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
Claim(s) 1-11 is/are method type claim. Claim(s) 12-19 is/are system type claim(s). Claim(s) 20 is/are product type claim(s). Therefore, claims 1-20 is/are directed to either a process, machine, manufacture or composition of matter.
Independent claim(s):
Step 2A Prong 1:
Regarding claim(s) 1, this/these claim(s) recite(s) encoding the original input data into latent representations,
Regarding claim(s) 12 this/these claim(s) recite(s) decode the synthetic representations, and
Regarding claim(s) 20, this/these claim(s) recite(s) encoding the original input data into latent representations.
The above limitations appear to be practically implementable in the human mind and is understood to be a recitation of a mental process – a user can mentally encode and decode information.
Step 2A Prong 2:
Regarding claim(s) 1, 12 and 20 this judicial exception is not integrated into a practical application.
Additional elements:
Claim 12, has been interpreted above as, "generator" and “decoder” are implemented as hardware/ processor(s) executing software modules to perform the instructions of the modules.
Regarding claim(s) 12 and 20, this/these claim(s) recite(s) processor and memory to perform the steps of abstract idea (mere instructions stored in a generic memory component to apply the exception using a generic computer component).
Regarding claim(s) 1, this/these claim(s) further recite(s)
receiving original input data (Adding insignificant extra-solution activity (receiving information) to the judicial exception - see MPEP 2106.05(g). Examiner’s note: the obtaining is recited at a high level of generality and could constitute mere receiving of transmitted information);
training an encoder-decoder model using the original input data, the encoder-decoder model comprising an encoder and a decoder (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training an encoder-decoder model comprising an encoder and a decoder, with previously determined data);
training a generative adversarial network (GAN) framework, including a generator and a discriminator, based on the latent representations (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training generative adversarial network (GAN) framework including a generator and a discriminator, with previously determined data).
Regarding claim(s) 12, this/these claim(s) further recite(s)
an encoder-decoder model comprising an encoder and a decoder; (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of application of an encoder-decoder model comprising an encoder and a decoder, with previously determined data); and
a generative adversarial network (GAN), comprising a generator and a discriminator, wherein the GAN is trained using latent representations from a training of the encoder-decoder 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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training a generative adversarial network (GAN), comprising a generator and a discriminator, with previously determined data); and
wherein in generating the synthetic data (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of application of the generator to generate synthetic data, with previously determined data),
the generator is configured to receive random sample vectors (Adding insignificant extra-solution activity (receiving information) to the judicial exception - see MPEP 2106.05(g). Examiner’s note: the obtaining is recited at a high level of generality and could constitute mere receiving of transmitted information).
Regarding claim(s) 20 this/these claim(s) further recite(s)
training an encoder-decoder model using the original input data, the encoder-decoder model comprising an encoder and a decoder (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training an encoder-decoder model comprising an encoder and a decoder, with previously determined data);,
training a generative adversarial network (GAN) framework, including a generator and a discriminator, based on the latent representations (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training generative adversarial network (GAN) framework including a generator and a discriminator, with previously determined data).
The additional element(s) as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above.
Therefore, the claim(s) is/are directed to an abstract idea.
Step 2B:
Regarding claim(s) 1, 12 and 20, this/these claim(s) do/does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Additional elements:
Claim 12, has been interpreted above as, "generator" and “decoder” are implemented as hardware/ processor(s) executing software modules to perform the instructions of the modules.
Regarding claim(s) 12 and 20, this/these claim(s) recite(s) processor and memory to perform the steps of abstract idea (mere instructions stored in a generic memory component to apply the exception using a generic computer component).
Regarding claim(s) 1, this/these claim(s) further recite(s)
receiving original input data (Adding insignificant extra-solution activity (receiving information) to the judicial exception - see MPEP 2106.05(g). Examiner’s note: the obtaining is recited at a high level of generality and could constitute mere receiving of transmitted information, this insignificant extra solution activity is well understood routine and conventional activity, see Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362);
training an encoder-decoder model using the original input data, the encoder-decoder model comprising an encoder and a decoder (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training an encoder-decoder model comprising an encoder and a decoder, with previously determined data);
training a generative adversarial network (GAN) framework, including a generator and a discriminator, based on the latent representations (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training generative adversarial network (GAN) framework including a generator and a discriminator, with previously determined data).
Regarding claim(s) 12, this/these claim(s) further recite(s)
an encoder-decoder model comprising an encoder and a decoder; (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of application of an encoder-decoder model comprising an encoder and a decoder, with previously determined data); and
a generative adversarial network (GAN), comprising a generator and a discriminator, wherein the GAN is trained using latent representations from a training of the encoder-decoder 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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training a generative adversarial network (GAN), comprising a generator and a discriminator, with previously determined data); and
wherein in generating the synthetic data (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of application of the generator to generate synthetic data, with previously determined data),
the generator is configured to receive random sample vectors (Adding insignificant extra-solution activity (receiving information) to the judicial exception - see MPEP 2106.05(g). Examiner’s note: the obtaining is recited at a high level of generality and could constitute mere receiving of transmitted information, this insignificant extra solution activity is well understood routine and conventional activity, see Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362).
Regarding claim(s) 20 this/these claim(s) further recite(s)
training an encoder-decoder model using the original input data, the encoder-decoder model comprising an encoder and a decoder (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training an encoder-decoder model comprising an encoder and a decoder, with previously determined data);,
training a generative adversarial network (GAN) framework, including a generator and a discriminator, based on the latent representations (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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training generative adversarial network (GAN) framework including a generator and a discriminator, with previously determined data).
The additional element(s) as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above.
Therefore, the claim(s) is/are not patent eligible.
Step 2A Prong 1, Dependent claims:
Regarding claim(s) 3, this/these claim(s) recite(s)
sampling, random vectors; generating, synthetic embeddings from the random vectors; and using, the synthetic embeddings to generate synthetic temporal and categorical data,
Regarding claim(s) 5 and 15, this/these claim(s) recite(s) generating missing patterns representing missing features of the original input data,
Regarding claim(s) 6 and 16, this/these claim(s) recite(s) generating original encoder states using the trained encoder, original input data, and the missing patterns,
Regarding claim(s) 8 and 17, this/these claim(s) recite(s) transforming categorical data into one-hot encoded data, transforming the one-hot encoded data into categorical embeddings,
Regarding claim(s) 13, this/these claim(s) recite(s) the decoder is configured to use the synthetic representations to generate synthetic temporal and categorical data, and
Regarding claim(s) 14, this/these claim(s) recite(s) encoding the original input data into the latent representations.
The above limitations appear to be practically implementable in the human mind and is understood to be a recitation of a mental process.
Regarding claim(s) 7, this/these claim(s) recite(s) wherein training the encoder-decoder model further comprises stochastic normalization for numerical features,
Regarding claim(s) 10 and 19, this/these claim(s) recite(s) wherein the encoder-decoder model is trained using reconstruction loss, and the GAN framework is trained using adversarial loss, and
Regarding claim(s) 11, this/these claim(s) recite(s) reconstruction loss uses mean square error for temporal features, measurement time, and static features
The above limitations appear to be a recitation of a mental process and math.
Step 2A Prong 2, Dependent claims:
Regarding claim(s) 2, this/these claim(s) recite(s) generating synthetic data using the trained generator and the trained decoder
Regarding claim(s) 3, this/these claim(s) recite(s) by the generator, by the decoder,
Regarding claim(s) 8 and 17, this/these claim(s) recite(s) training a temporal categorical encoder and a temporal categorical decoder, and
Regarding claim(s) 14, this/these claim(s) recite(s) encoder-decoder model is trained using original input data, ... latent representations ... are provided to the GAN.
The above limitations appear to add 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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training or using encoder-decoder model and/or GAN framework);
Regarding claim(s) 4, this/these claim(s) recite(s) wherein the original input data comprises one or more of static numeric features, static categorical features, temporal numeric features, temporal categorical features, or measurement time,
Regarding claim(s) 9 and 18, this/these claim(s) recite(s) wherein the original input data comprises heterogenous time- series data
These limitations appear to be directed to the specification of data to be used, and is understood to be generally linking the use of the judicial exception to a particular technological environment or field of use.
Step 2B, Dependent claims:
Regarding claim(s) 2, this/these claim(s) recite(s) generating synthetic data using the trained generator and the trained decoder
Regarding claim(s) 3, this/these claim(s) recite(s) by the generator, by the decoder,
Regarding claim(s) 8 and 17, this/these claim(s) recite(s) training a temporal categorical encoder and a temporal categorical decoder, and
Regarding claim(s) 14, this/these claim(s) recite(s) encoder-decoder model is trained using original input data, ... latent representations ... are provided to the GAN.
The above limitations appear to add 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 - see MPEP 2106.05(f) – Examiner’s note: high level recitation of training or using encoder-decoder model and/or GAN framework);
Regarding claim(s) 4, this/these claim(s) recite(s) wherein the original input data comprises one or more of static numeric features, static categorical features, temporal numeric features, temporal categorical features, or measurement time,
Regarding claim(s) 9 and 18, this/these claim(s) recite(s) wherein the original input data comprises heterogenous time- series data
These limitations appear to be directed to the specification of data to be used, and is understood to be generally linking the use of the judicial exception to a particular technological environment or field of use.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1, 2, 4, 10, 12, 14, 19, 20, **** is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Stojevic (US 20210081804 A1).
Regarding claim 1, Stojevic teaches a method, comprising (Stojevic [133, 274] method for a system processor to execute instructions stored in memory to generate synthetic data):
receiving original input data (Stojevic [147, 148] input data received);
training an encoder-decoder model using the original input data, the encoder-decoder model comprising an encoder and a decoder (Stojevic [26, 27, 147] autoencoder (which has an encoder and decoder) may be trained using input data);
encoding the original input data into latent representations (Stojevic [128] latent space encoding generated for input data); and
training a generative adversarial network (GAN) framework, including a generator and a discriminator, based on the latent representations (Stojevic Fig. 18, [147, 163, 164, 166] GAN framework includes generator and a discriminator, GAN framework may be trained using latent space encoding).
Regarding claim 2, Stojevic teaches the invention as claimed in claim 1 above.
Stojevic further teaches generating synthetic data using the trained generator and the trained decoder (Stojevic [118, 150, 151] trained generator and trained decoder used to generate synthetic data).
Regarding claim 4, Stojevic teaches the invention as claimed in claim 1 above.
Stojevic further teaches wherein the original input data comprises ... static categorical features (Stojevic [160, 161] output data may include classification, output data may reconstruct input data (so the same)).
Regarding claim 10, Stojevic teaches the invention as claimed in claim 1 above.
Stojevic further teaches wherein the encoder-decoder model is trained using reconstruction loss, and the GAN framework is trained using adversarial loss (Stojevic [140, 164-166] various cost functions (loss) are optimized while training the framework, including the difference between autoencoder output and input (reconstruction loss) and whether discriminator can real and fake samples correctly (adversarial loss)).
Regarding claim 12, Stojevic teaches a system for generating synthetic data, comprising generator and decoder (Stojevic [133, 274] method for a system processor to execute instructions stored in memory to generate synthetic data):
an encoder-decoder model comprising an encoder and a decoder (Stojevic [26, 27, 147] autoencoder (which has an encoder and decoder) may be trained using input data);
a generative adversarial network (GAN), comprising a generator and a discriminator, wherein the GAN is trained using latent representations from a training of the encoder-decoder model (Stojevic Fig. 18, [147, 163, 164, 166] GAN framework includes generator and a discriminator, GAN framework may be trained using latent space encoding, Stojevic [128] trained autoencoder may generate latent space encoding generated for input data); and
wherein in generating the synthetic data, the generator is configured to receive random sample vectors and generate synthetic representations, and the decoder is configured to decode the synthetic representations (Stojevic [14, 150, 462] data may be generated based on sampling which may include sampling random vector representation (tensor), Stojevic [110] generator may embed tensors, Stojevic [125, 128, 150, 151, 160, 169] decoder may use embeddings to generate classification (categorial) data).
Regarding claim 14, Stojevic teaches the invention as claimed in claim 12 above.
Stojevic further teaches wherein the encoder-decoder model is trained using original input data, the training including encoding the original input data into the latent representations that are provided to the GAN (Stojevic [26, 27, 147] autoencoder (which has an encoder and decoder) may be trained using input data, Stojevic [128] latent space encoding generated for input data, Stojevic Fig. 18, [147, 163, 164, 166] GAN framework includes generator and a discriminator, GAN framework may be trained using latent space encoding)
Regarding claim 19, Stojevic teaches the invention as claimed in claim 12 above.
Stojevic further teaches wherein the encoder-decoder model is trained using reconstruction loss, and the GAN framework is trained using adversarial loss (Stojevic [140, 164-166] various cost functions (loss) are optimized while training the framework, including the difference between autoencoder output and input (reconstruction loss) and whether discriminator can real and fake samples correctly (adversarial loss)).
Claim 20 is directed towards a medium storing instructions similar in scope to the instructions performed by the method of claim 1, and is rejected under the same rationale.
Stojevic further teaches a non-transitory computer-readable medium storing instructions executable by one or more processors to perform a method ... (Stojevic [133, 274] method for a system processor to execute instructions stored in memory to generate synthetic data):
the training including encoding the original input data into latent representations (Stojevic [26, 27, 147] autoencoder (which has an encoder and decoder) may be trained using encoding in latent space).
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.
Claim(s) 3, 9, 13, 18, is/are rejected under 35 U.S.C. 103 as being unpatentable over Stojevic (US 20210081804 A1), in view of Ryan (US 20200387797 A1).
Regarding claim 3, Stojevic teaches the invention as claimed in claim 2 above.
Stojevic further teaches wherein generating the synthetic data comprises: sampling, by the generator, random vectors (Stojevic [14, 150, 462] data may be generated based on sampling which may include sampling random vector representation (tensor));
generating, by the generator, synthetic embeddings from the random vectors (Stojevic [110] generator may embed tensors); and
using, by the decoder, the synthetic embeddings to generate synthetic...categorical data (Stojevic [125, 128, 150, 151, 160, 169] decoder may use embeddings to generate classification (categorial) data).
Stojevic does not specifically teach using, by the decoder, the synthetic embeddings to generate synthetic temporal data
However Ryan teaches using, by the decoder, the synthetic embeddings to generate synthetic temporal data (Ryan [128, 129, 132, 158, 201, 203, 215] synthetic data may be generated by decoder based on synthetic embeddings, data generated may include temporal data).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Ryan of using, by the decoder, the synthetic embeddings to generate synthetic temporal data, into the invention suggested by Stojevic; since both inventions are directed towards training a GAN framework to generate synthetic outputs, and incorporating the teaching of Ryan into the invention suggested by Stojevic would provide the added advantage of allowing temporal data to be generated thereby allowing analysis of data trends to detect anomalies, and the combination would perform with a reasonable expectation of success (Ryan [128, 129, 132, 158, 201, 203, 215]).
Regarding claims 9 and 18, Stojevic teach(es) the invention as claimed in claims 1 and 14 above.
Stojevic does not specifically teach wherein the original input data comprises heterogenous time- series data
However Ryan teaches wherein the original input data comprises heterogenous time- series data (Ryan Abstract [80, 129, 130, 160] input may be time series which may abrupt or cyclic changes, anomalies can be determined based on data trends).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Ryan of wherein the original input data comprises heterogenous time- series data, into the invention suggested by Stojevic; since both inventions are directed towards training a GAN framework to generate synthetic outputs, and incorporating the teaching of Ryan into the invention suggested by Stojevic would provide the added advantage of allowing anomalies to be determined based on data trends, and the combination would perform with a reasonable expectation of success
(Ryan Abstract [80, 129, 130, 160]).
Regarding claim 13, Stojevic teach(es) the invention as claimed in claim 12 above.
Stojevic further teaches wherein in decoding the synthetic representations, the decoder is configured to use the synthetic representations to generate synthetic ... categorical data (Stojevic [125, 128, 150, 151, 160, 169] decoder may use embeddings to generate classification (categorial) data).
Stojevic does not specifically teach using, by the decoder, the synthetic embeddings to generate synthetic temporal data
However Ryan teaches the decoder is configured to use the synthetic representations to generate synthetic temporal ... data (Ryan [128, 129, 132, 158, 201, 203, 215] synthetic data may be generated by decoder based on synthetic embeddings, data generated may include temporal data).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Ryan of the decoder is configured to use the synthetic representations to generate synthetic temporal ... data, into the invention suggested by Stojevic; since both inventions are directed towards training a GAN framework to generate synthetic outputs, and incorporating the teaching of Ryan into the invention suggested by Stojevic would provide the added advantage of allowing temporal data to be generated thereby allowing analysis of data trends to detect anomalies, and the combination would perform with a reasonable expectation of success (Ryan [128, 129, 132, 158, 201, 203, 215]).
Claim(s) 5, 6, 15, 16, is/are rejected under 35 U.S.C. 103 as being unpatentable over Stojevic (US 20210081804 A1), in view of Brauer (US 20220375051 A1).
Regarding claims 5 and 15, Stojevic teach(es) the invention as claimed in claims 1 and 14 above. Stojevic does not specifically teach generating missing patterns representing missing features of the original input data
However Brauer teaches generating missing patterns representing missing features of the original input data (Brauer [86] missing patterned features may be identified, to successfully generate a trained deep generative model even if the data used for training is missing some of the patterned features that appear in or have an impact on data).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Brauer of generating missing patterns representing missing features of the original input data, into the invention suggested by Stojevic; since both inventions are directed towards generative models, and incorporating the teaching of Brauer into the invention suggested by Stojevic would provide the added advantage of successfully generate a trained deep generative model even if the data used for training is missing some of the patterned features that appear in or have an impact on data, and the combination would perform with a reasonable expectation of success
(Brauer [86]).
Regarding claims 6 and 16 Stojevic and Brauer teach(es) the invention as claimed in claims 5 and 15 above.
Stojevic does not specifically teach generating original encoder states using the trained encoder, original input data, and the missing patterns
However Brauer teaches generating original encoder states using the trained encoder, original input data, and the missing patterns (Brauer [66, 72, 77, 86, 95] missing patterned features may be identified, trained generator encoder may reconstruct encoding based on input data and identified missing patterns).
Claim(s) 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stojevic (US 20210081804 A1), in view of Rajanna (US 20240420316 A1).
Regarding claim 7, Stojevic teach(es) the invention as claimed in claim 1 above.
Stojevic does not specifically teach training the encoder-decoder model further comprises stochastic normalization for numerical features
However Rajanna teaches training the encoder-decoder model further comprises stochastic normalization for numerical features (Rajanna [4, 17, 73, 101, 104, 108, 127] encoder may be in GAN system, encoder may be trained using hash codes which are stochastic normalization (random distribution) for numerical features).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Rajanna of training the encoder-decoder model further comprises stochastic normalization for numerical features, into the invention suggested by Stojevic; since both inventions are directed towards GAN systems, and incorporating the teaching of Rajanna into the invention suggested by Stojevic would provide the added advantage of introducing diversity to the training by using stochastic normalization, and the combination would perform with a reasonable expectation of success
(Rajanna [4, 17, 73, 101, 104, 108, 127]).
Claim(s) 8 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stojevic (US 20210081804 A1), in view of Ryan (US 20200387797 A1) and Triendl (US 20220383992 A1).
Regarding claim 8 and 17 Stojevic teach(es) the invention as claimed in claims 1 and 14 above. Stojevic does not specifically teach wherein training the encoder-decoder model comprises: transforming categorical data into one-hot encoded data; training a temporal categorical encoder and a temporal categorical decoder; and transforming the one-hot encoded data into categorical embeddings
However Ryan teaches training a temporal categorical encoder and a temporal categorical decoder (Ryan [163, 201-203, 215] encoder and decoder may be trained and used for time-series classifications, data input may include temporal data, Ryan Abstract [80, 129, 130, 160] input may be time series which may abrupt or cyclic changes, anomalies can be determined based on data trends).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Ryan of training a temporal categorical encoder and a temporal categorical decoder, into the invention suggested by Stojevic; since both inventions are directed towards training a GAN framework to generate synthetic outputs, and incorporating the teaching of Ryan into the invention suggested by Stojevic would provide the added advantage of allowing temporal data to be generated thereby allowing analysis of data trends to detect anomalies, and the combination would perform with a reasonable expectation of success (Ryan Abstract [163, 201-203, 215, 80, 129, 130, 160]).
Stojevic and Ryan does not specifically teach wherein training the encoder-decoder model comprises: transforming categorical data into one-hot encoded data; and transforming the one-hot encoded data into categorical embeddings.
However Triendl teaches wherein training the encoder-decoder model comprises: transforming categorical data into one-hot encoded data; and transforming the one-hot encoded data into categorical embeddings (Triendl [187, 306, 307] GAN used, atom types and bond types (categorical) may be on-hot encoded and then converted to feature vectors for sparse representation).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Triendl of wherein training the encoder-decoder model comprises: transforming categorical data into one-hot encoded data; and transforming the one-hot encoded data into categorical embeddings, into the invention suggested by Stojevic and Ryan; since both inventions are directed towards GANs, and incorporating the teaching of Triendl into the invention suggested by Stojevic and Ryan would provide the added advantage of sparse representation of input data, and the combination would perform with a reasonable expectation of success (Triendl [187, 306, 307]).
Claim(s) 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Stojevic (US 20210081804 A1), in view of Salle (US 20210295105 A1) and Ryan (US 20200387797 A1).
Regarding claim 11 Stojevic teach(es) the invention as claimed in claim 10 above.
Stojevic does not specifically teach wherein reconstruction loss uses mean square error for temporal features, measurement time, and static features
However Salle teaches wherein reconstruction loss uses mean square error for temporal features, ... and static features (Salle [20, 29, 37, 38] statistical methods can be used to determine reconstruction loss using dynamic and static features, therefore learning both parameters that are invariant to a given transformation and parameters that characterize the transformation).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Salle of wherein reconstruction loss uses mean square error for temporal features, ... and static features, into the invention suggested by Stojevic; since both inventions are directed towards using reconstruction loss in a GAN system, and incorporating the teaching of Salle into the invention suggested by Stojevic would provide the added advantage of learning both parameters that are invariant to a given transformation and parameters that characterize the transformation, and the combination would perform with a reasonable expectation of success
(Salle [20, 29, 37, 38]).
Stojevic and Salle does not specifically teach measurement time
However Ryan teaches wherein reconstruction ... includes ... measurement time (Ryan 132] time stamp (measurement time) is beneficial in anomaly detection for temporal data).
It would have been obvious to one of an ordinary skill in the art before the effective filing date of the claimed invention, to have incorporated the concept taught by Ryan of wherein reconstruction ... includes ... measurement time, into the invention suggested by Stojevic;
so that the reconstruction loss ...that...uses mean square error as taught by Stojevic and Salle, includes reconstruction loss for
reconstruction ... includes ... measurement time as taught by Ryan,
since both inventions are directed towards training a GAN framework to generate synthetic outputs, and incorporating the teaching of Ryan into the invention suggested by Stojevic and Salle would provide the added advantage of being beneficial in anomaly detection for temporal data, and the combination would perform with a reasonable expectation of success
(Ryan 132]).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Hagi (US 20220180527 A1) discloses a GAN topology to generate simulated data using latent representations of input data.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SANCHITA ROY whose telephone number is (571)272-5310. The examiner can normally be reached Monday-Friday 12-8.
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, Usmaan Saeed can be reached at (571) 272-4046. 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.
SANCHITA ROY
Primary Examiner
Art Unit 2146
/SANCHITA ROY/Primary Examiner, Art Unit 2146
Prosecution Insights