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 .
This communication is responsive to the application filed 07/11/2024.
Claims 1-20 are pending with claims 1, 8, and 15 as independent claims.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 10/04/2024 was filed after the mailing date of the application on 07/11/2024. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Objections
Claims 18-20 are objected to because of the following informalities: the claims appear to refer back to independent claim 15 not claim 16 because claims refer back to limitations in independent claim 15. Appropriate correction is required.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 1-20 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-20 of U.S. Patent No. 11,645,543. Although the claims at issue are not identical, they are not patentably distinct from each other.
The specification recites in [0089] “transaction service provider system 102 may provide a training dataset as a second input to the discriminator network. In some non-limiting embodiments, such (second) input may be provided (e.g., by transaction service provider system 102) to the discriminator network during a training procedure.”
Bold text indicates differences between Application and Patent.
Underlined text indicates content exist only in the patent.
Italic text is to indicate context.
Application
Patent
1. A computer-implemented method, comprising:
providing, with at least one processor, an input to a generator network of a generative adversarial network (GAN),
wherein the input comprises a first dataset, and wherein providing the input to the generator network of the GAN comprises:
1. A computer-implemented method for generating a machine learning model to classify an account based on merchant activation comprising:
providing, during a training procedure and with at least one processor, an input to a generator network of a generative adversarial network (GAN)),
wherein providing, during the training procedure, the input to the generator network of the GAN comprises:
providing randomly generated data as an input to a tree based machine learning model and a Neural machine learning model; and
providing an output of the tree based machine learning model and the Neural machine learning model as the input to the generator network of the GAN; [1]
providing randomly generated data as an input to an XGBoost machine learning model and a Neural Collaborative Filtering machine learning model; and
providing an output of the XGBoost machine learning model and the Neural Collaborative Filtering machine learning model as the input to the generator network of the GAN;
generating, with at least one processor, an output of the generator network based on the input to the generator network of the GAN, wherein the output comprises a generated dataset, wherein the generated dataset comprises a first plurality of sets of values for each of a plurality of features and the first plurality of sets of values for each of the plurality of features comprises data associated with a first plurality of payment transactions,
wherein the plurality of features of the first dataset is the same as the plurality of features of the generated dataset, and wherein the plurality of features comprises: [2]
a feature associated with a time interval of a payment transaction;
a feature associated with a market segment of a merchant involved in a payment transaction;
a feature associated with a transaction amount of a payment transaction; and
a feature associated with a total amount of a plurality of payment transactions conducted within a time interval;
generating, with at least one processor, an output of the generator network based on the input to the generator network of the GAN, wherein the output comprises a generated dataset, wherein the generated dataset comprises a first plurality of sets of values for each of a plurality of features and the first plurality of sets of values for each of the plurality of features comprises data associated with a first plurality of payment transactions
conducted using a plurality of accounts and involving a plurality of merchants,
wherein the plurality of features of the training dataset is the same as the plurality of features of the generated dataset, and wherein the plurality of features comprises:
a feature associated with a time interval of a payment transaction,
a feature associated with a market segment of a merchant involved in a payment transaction,
a feature associated with a transaction amount of a payment transaction,
a feature associated with a total amount of a plurality of payment transactions conducted within a time interval, and
a feature associated with a total amount of a plurality of card-present payment transactions conducted within a time interval;
providing, with at least one processor, the output of the generator network as an input to a discriminator network of the GAN;
providing, during the training procedure and with at least one processor, the output of the generator network as an input to a discriminator network of the GAN;
providing, with at least one processor, a second dataset as an input to the discriminator network of the GAN, wherein the second dataset comprises a second plurality of sets of values for each of the plurality of features and the second plurality of sets of values for each of the plurality of features comprises data associated with a second plurality of payment transactions; and
providing, during the training procedure and with at least one processor, a training dataset as an input to the discriminator network of the GAN, wherein the training dataset comprises a second plurality of sets of values for each of the plurality of features and the second plurality of sets of values for each of the plurality of features comprises data associated with a second plurality of payment transactions
conducted using the plurality of accounts and involving the plurality of merchants;
updating, with at least one processor, the discriminator network of the GAN based on a second output of the discriminator network of the GAN having a label that indicates whether a selected account of a plurality of accounts is going to conduct a first payment transaction with a selected entity of a plurality of entities. [3]
updating, during the training procedure and with at least one processor, the discriminator network of the GAN based on a second output of the discriminator network of the GAN having a label that indicates whether a selected account of the plurality of accounts is going to conduct a first payment transaction with a selected merchant of the plurality of merchants.
[1] The tree based machine learning model may be an XGBoost machine learning model because the XGBoost machine learning model is actually a type of free-based machine learning model. The neural machine learning model may be a neural collaborative filtering machine learning model.
[2] The first dataset may be training dataset, which may be based on fake training data that is input to the generator to output generated dataset as an input to the discriminator. The second dataset may be also training dataset (indicated by the specification in [0089]), based on real training data, which is also input to the discriminator.
[3] The limitation entity may include a merchant.
With regard to [2], assume that the features of the training dataset indicated by the patent is based on real training data, then the features of the first dataset, which is the same as the plurality of features of the generated dataset that is input to the discriminator, may be based on fake training data that is input to the discriminator. Accordingly, it would be obvious that the first dataset requires to have features of payment transactions, based on fake training data, similar to the training dataset (second dataset), based on real training data so that the discriminator can be trained to accurately identifies fake or suspicious transactions.
Application
Patent
2. The computer-implemented method of claim 1, wherein the generator network of the GAN comprises at least one of the following:
a first dense layer comprising a rectified linear unit (Relu) function with 16 nodes;
a second dense layer comprising a Relu function with 32 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 64 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 128 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a Relu function with 182 nodes, wherein the fifth dense layer is fully connected to the fourth dense layer; or
any combination thereof.
2. The computer-implemented method of claim 1, wherein the generator network of the GAN comprises:
a first dense layer comprising a rectified linear unit (Relu) function with 16 nodes;
a second dense layer comprising a Relu function with 32 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 64 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 128 nodes, wherein the fourth dense layer is fully connected to the third dense layer; and
a fifth dense layer comprising a Relu function with 182 nodes, wherein the fifth dense layer is fully connected to the fourth dense layer.
Application
Patent
3. The computer-implemented method of claim 2, further comprising:
during a training procedure, implementing a dropout at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN; and
during the training procedure, implementing a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN.
3. The computer-implemented method of claim 2, further comprising:
during the training procedure, implementing a dropout of 30% at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN; and
during the training procedure, implementing a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN.
Application
Patent
4. The computer-implemented method of claim 1, wherein the discriminator network of the GAN comprises at least one of the following:
a first dense layer comprising a rectified linear unit (Relu) function with 128 nodes;
a second dense layer comprising a Relu function with 64 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 32 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 16 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a sigmoid function with 1 node, wherein the fifth dense layer is fully connected to the fourth dense layer;
a sixth dense layer comprising a sigmoid function with 1 node, wherein the sixth dense layer is fully connected to the fourth dense layer; or any combination thereof.
4. The computer-implemented method of claim 1, wherein the discriminator network of the GAN comprises:
a first dense layer comprising a Relu function with 128 nodes;
a second dense layer comprising a Relu function with 64 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 32 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 16 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a sigmoid function with 1 node, wherein the fifth dense layer is fully connected to the fourth dense layer; and
a sixth dense layer comprising a sigmoid function with 1 node, wherein the sixth dense layer is fully connected to the fourth dense layer.
Application
Patent
5. The computer-implemented method of claim 4, further comprising:
during a training procedure, implementing a dropout at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN; and
during the training procedure, implementing a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN.
5. The computer-implemented method of claim 4, further comprising:
during the training procedure, implementing a dropout of 30% at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN; and
during the training procedure, implementing a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN.
Application
Patent
6. The computer-implemented method of claim 1, further comprising:
providing an input to the discriminator network of the GAN and obtaining an output that indicates whether an account is going to conduct a first payment transaction with a merchant of the plurality of merchants.
6. The computer-implemented method of claim 1, further comprising:
providing an input to the discriminator network of the GAN and obtaining an output that indicates whether an account is going to conduct a first payment transaction with a merchant of the plurality of merchants.
Application
Patent
7. The computer-implemented method of claim 1, further comprising:
optimizing the discriminator network of the GAN based on a formula, wherein the formula includes a value of Recall for the discriminator network and the formula is defined as:
Recall= TP / (TP + FN),
wherein TP is a number of true positive predictions based on an output of the discriminator network corresponding to a ground truth label of a set of values of a plurality of features, and
wherein FN is a number of false negative predictions based on an output of the discriminator network not corresponding to a ground truth label of a set of values of a plurality of features.
8. The computer-implemented method of claim 1, further comprising:
during the training procedure, optimizing the discriminator network of the GAN based on a formula, wherein the formula includes a value of Recall for the discriminator network and the formula is defined as:
Recall= TP / (TP + FN);
wherein TP is a number of true positive predictions based on an output of the discriminator network corresponding to a ground truth label of a set of values of a plurality of features; and
wherein FN is a number of false negative predictions based on an output of the discriminator network not corresponding to a ground truth label of a set of values of a plurality of features.
Application
Patent
8. A system, comprising:
at least one processor programmed or configured to:
provide an input to a generator network of a generative adversarial network (GAN), wherein the input comprises a first dataset, and wherein, when providing the input to the generator network of the GAN, the at least one processor is programmed or configured to:
9. A system for generating a machine learning model to classify an account based on merchant activation, comprising:
at least one processor programmed or configured to:
provide, during a training procedure, an input to a generator network of a generative adversarial network (GAN), wherein, when providing, during the training procedure, the input to the generator network of the GAN, the at least one processor is programmed or configured to:
provide randomly generated data as an input to a tree based machine learning model and a Neural network machine
learning model; and
provide an output of the tree based machine learning model and the Neural network machine learning model as the input
to the generator network of the GAN; [1]
provide randomly generated data as an input to an XGBoost machine learning model and a Neural Collaborative Filtering machine learning model; and
provide an output of the XGBoost machine learning model and the Neural Collaborative Filtering machine learning model as the input to the generator network of the GAN;
generate an output of the generator network based on the input to the generator network of the GAN, wherein the output comprises a generated dataset, wherein the generated dataset comprises a first plurality of sets of values for each of a plurality of features and the first plurality of sets of values for each of the plurality of features comprises data associated with a first plurality of payment transactions,
wherein the plurality of features of the first dataset is the same as the plurality of features of the generated dataset, and wherein the plurality of features comprises:
a feature associated with a time interval of a payment transaction;
a feature associated with a market segment of a merchant involved in a payment transaction;
a feature associated with a transaction amount of a payment transaction; and
a feature associated with a total amount of a plurality of payment transactions conducted within a time interval;
generate an output of the generator network based on the input to the generator network of the GAN, wherein the output comprises a generated dataset, wherein the generated dataset comprises a first plurality of sets of values for each of a plurality of features and the first plurality of sets of values for each of the plurality of features comprises data associated with a first plurality of payment transactions
conducted using a plurality of accounts and involving a plurality of merchants;
provide the output of the generator network as an input to a discriminator network of the GAN;
provide, during the training procedure, the output of the generator network as an input to a discriminator network of the GAN;
provide a second dataset as an input to the discriminator network of the GAN, wherein the second dataset comprises a second plurality of sets of values for each of the plurality of features and the second plurality of sets of values for each of the plurality of features comprises data associated with a second plurality of payment transactions; and [2]
provide, during the training procedure, a training dataset as an input to the discriminator network of the GAN, wherein the training dataset comprises a second plurality of sets of values for each of the plurality of features and the second plurality of sets of values for each of the plurality of features comprises data associated with a second plurality of payment transactions
conducted using the plurality of accounts and involving the plurality of merchants,
wherein the plurality of features of the training dataset is the same as the plurality of features of the generated dataset, and wherein the plurality of features comprises:
a feature associated with a time interval of a payment transaction, a feature associated with a market segment of a merchant involved in a payment transaction,
a feature associated with a transaction amount of a payment transaction,
a feature associated with a total amount of a plurality of payment transactions conducted within a time interval, and
a feature associated with a total amount of a plurality of card-present payment transactions conducted within a time interval;
update the discriminator network of the GAN based on a second output of the discriminator network of the GAN having a label that indicates whether a selected account of a plurality of accounts is going to conduct a first payment transaction with a selected entity of a plurality of entities. [3]
update, during the training procedure, the discriminator network of the GAN based on a second output of the discriminator network of the GAN having a label that indicates whether a selected account of the plurality of accounts is going to conduct a first payment transaction with a selected merchant of the plurality of merchants.
[1] The tree based machine learning model may be an XGBoost machine learning model because the XGBoost machine learning model is actually a type of free-based machine learning model. The neural machine learning model may be a neural collaborative filtering machine learning model.
[2] The first dataset may be training dataset, which may be based on fake training data that is input to the generator to output generated dataset as an input to the discriminator. The second dataset may be also training dataset (indicated by the specification in [0089]), based on real training data, which is also input to the discriminator.
[3] The limitation entity may include a merchant.
With regard to [2], assume that the features of the training dataset indicated by the patent is based on real training data, then the features of the first dataset, which is the same as the plurality of features of the generated dataset that is input to the discriminator, may be based on fake training data that is input to the discriminator. Accordingly, it would be obvious that the first dataset requires to have features of payment transactions, based on fake training data, similar to the training dataset (second dataset), based on real training data so that the discriminator can be trained to accurately identifies fake or suspicious transactions.
Application
Patent
9. The system of claim 8, wherein the generator network of the GAN comprises at least one of the following:
a first dense layer comprising a rectified linear unit (Relu) function with 16 nodes;
a second dense layer comprising a Relu function with 32 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 64 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 128 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a Relu function with 182 nodes, wherein the fifth dense layer is fully connected to the fourth dense layer; or
any combination thereof.
2. The computer-implemented method of claim 1, wherein the generator network of the GAN comprises:
a first dense layer comprising a rectified linear unit (ReLu) function with 16 nodes;
a second dense layer comprising a ReLu function with 32 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a ReLu function with 64 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a ReLu function with 128 nodes, wherein the fourth dense layer is fully connected to the third dense layer; and
a fifth dense layer comprising a ReLu function with 182 nodes, wherein the fifth dense layer is fully connected to the fourth dense layer.
Application
Patent
10. The system of claim 9, wherein the at least one processor is further programmed or configured to:
during a training procedure, implement a dropout at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN; and
during the training procedure, implement a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN.
10. The system of claim 9, wherein the at least one processor is further programmed or configured to:
during the training procedure, implement a dropout of 30% at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN; and
during the training procedure, implement a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, the fourth dense layer, and the fifth dense layer of the generator network of the GAN.
Application
Patent
11. The system of claim 8, wherein the discriminator network of the GAN comprises at least one of the following:
a first dense layer comprising a rectified linear unit (Relu) function with 128 nodes;
a second dense layer comprising a Relu function with 64 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 32 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 16 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a sigmoid function with 1 node, wherein the fifth dense layer is fully connected to the fourth dense layer;
a sixth dense layer comprising a sigmoid function with 1 node, wherein the sixth dense layer is fully connected to the fourth dense layer; or any combination thereof.
11. The system of claim 9, wherein the discriminator network of the GAN
comprises:
a first dense layer comprising a Relu function with 128 nodes;
a second dense layer comprising a Relu function with 64 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 32 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 16 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a sigmoid function with 1 node, wherein the fifth dense layer is fully connected to the fourth dense layer; and
a sixth dense layer comprising a sigmoid function with 1 node, wherein the sixth dense layer is fully connected to the fourth dense layer.
Application
Patent
12. The system of claim 11, wherein the at least one processor is programmed or configured to:
during a training procedure, implement a dropout at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN; and
during the training procedure, implement a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN.
12. The system of claim 11, wherein the at least one processor is programmed or configured to:
during the training procedure, implement a dropout of 30% at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN; and
during the training procedure, implement a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN.
Application
Patent
13. The system of claim 8, wherein the at least one processor is further programmed or configured to:
provide an input to the discriminator network of the GAN and obtain an output that indicates whether an account is going to conduct a first payment transaction with a merchant of a plurality of merchants.
13. The system of claim 9, wherein the at least one processor is further programmed or configured to:
provide an input to the discriminator network of the GAN and obtain an output that indicates whether an account is going to conduct a first payment transaction with a merchant of the plurality of merchants.
Application
Patent
14. The system of claim 8, wherein the at least one processor is further programmed or configured to:
optimize the discriminator network of the GAN based on a formula, wherein the formula includes a value of Recall for the discriminator network and the formula is defined as:
Recall= TP / (TP + FN),
wherein TP is a number of true positive predictions based on an output of the discriminator network corresponding to a ground truth label of a set of values of a plurality of features, and
wherein FN is a number of false negative predictions based on an output of the discriminator network not corresponding to a ground truth label of a set of values of a plurality of features.
15. The system of claim 9, wherein the at least one processor is further programmed or configured to:
during the training procedure, optimize the discriminator network of the GAN based on a formula, wherein the formula includes a value of Recall for the discriminator network and the formula is defined as:
Recall= TP / (TP + FN);
wherein TP is a number of true positive predictions based on an output of the discriminator network corresponding to a ground truth label of a set of values of a plurality of features; and
wherein FN is a number of false negative predictions based on an output of the discriminator network not corresponding to a ground truth label of a set of values of a plurality of features.
Application
Patent
15. A computer program product comprising at least one non-transitory computer-readable medium including one or more instructions that, when executed by at least one processor, cause the at least one processor to:
16. A computer program product for generating a machine learning model to classify an account based on merchant activation comprising at least one non-transitory computer-readable medium including one or more instructions that, when executed by at least one processor, cause the at least one processor to:
provide an input to a generator network of a generative adversarial network (GAN), wherein the input comprises a first dataset, and wherein, the one or more instructions that cause the at least one processor to provide the input to the generator network of the GAN, cause the at least one processor to:
provide randomly generated data as an input to a tree based machine learning model and a Neural network machine learning model; and
provide an output of the tree based machine learning model and the Neural network machine learning model as the input to the generator network of the GAN; [1]
provide, during a training procedure, an input to a generator network of a generative adversarial network (GAN), wherein, the one or more instructions that cause the at least one processor to provide, during the training procedure, the input to the generator network of the GAN, cause the at least one processor to:
provide randomly generated data as an input to an XGBoost machine learning model and a Neural Collaborative Filtering machine learning model; and
provide an output of the XGBoost machine learning model and the Neural Collaborative Filtering machine learning model as the input to the generator network of the GAN;
generate an output of the generator network based on the input to the generator network of the GAN, wherein the output comprises a generated dataset, wherein the generated dataset comprises a first plurality of sets of values for each of a plurality of features and the first plurality of sets of values for each of the plurality of features comprises data associated with a first plurality of payment transactions,
wherein the plurality of features of the first dataset is the same as the plurality of features of the generated dataset, and wherein the plurality of features comprises:
a feature associated with a time interval of a payment transaction;
a feature associated with a market segment of a merchant involved in a payment transaction;
a feature associated with a transaction amount of a payment transaction; and
a feature associated with a total amount of a plurality of payment transactions conducted within a time interval;
generate an output of the generator network based on the input to the generator network of the GAN, wherein the output comprises a generated dataset, wherein the generated dataset comprises a first plurality of sets of values for each of a plurality of features and the first plurality of sets of values for each of the plurality of features comprises data associated with a first plurality of payment transactions
conducted using a plurality of accounts and involving a plurality of merchants;
wherein the plurality of features of the training dataset is the same as the plurality of features of the generated dataset, and wherein the plurality of features comprises:
a feature associated with a time interval of a payment transaction,
a feature associated with a market segment of a merchant involved in a payment transaction,
a feature associated with a transaction amount of a payment transaction,
a feature associated with a total amount of a plurality of payment transactions conducted within a time interval, and
a feature associated with a total amount of a plurality of card-present payment transactions conducted within a time interval;
provide the output of the generator network as an input to a discriminator network of the GAN;
provide, during the training procedure, the output of the generator network as an input to a discriminator network of the GAN;
provide a second dataset as an input to the discriminator network of the GAN, wherein the second dataset comprises a second plurality of sets of values for each of the plurality of features and the second plurality of sets of values for each of the plurality of features comprises data associated with a second plurality of payment transactions; and [2]
provide, during the training procedure, a training dataset as an input to the discriminator network of the GAN, wherein the training dataset comprises a second plurality of sets of values for each of the plurality of features and the second plurality of sets of values for each of the plurality of features comprises data associated with a second plurality of payment transactions
conducted using the plurality of accounts and involving the plurality of merchants
update the discriminator network of the GAN based on a second output of the discriminator network of the GAN having a label that indicates whether a selected account of a plurality of accounts is going to conduct a first payment transaction with a selected entity of a plurality of entities. [3]
update, during the training procedure, the discriminator network of the GAN based on a second output of the discriminator network of the GAN having a label that indicates whether a selected account of the plurality of accounts is going to conduct a first payment transaction with a selected merchant of the plurality of merchants.
[1] The tree based machine learning model may be an XGBoost machine learning model because the XGBoost machine learning model is actually a type of free-based machine learning model. The neural machine learning model may be a neural collaborative filtering machine learning model.
[2] The first dataset may be training dataset, which may be based on fake training data that is input to the generator to output generated dataset as an input to the discriminator. The second dataset may be also training dataset (indicated by the specification in [0089]), based on real training data, which is also input to the discriminator.
[3] The limitation entity may include a merchant.
With regard to [2], assume that the features of the training dataset indicated by the patent is based on real training data, then the features of the first dataset, which is the same as the plurality of features of the generated dataset that is input to the discriminator, may be based on fake training data that is input to the discriminator. Accordingly, it would be obvious that the first dataset requires to have features of payment transactions, based on fake training data, similar to the training dataset (second dataset), based on real training data so that the discriminator can be trained to accurately identifies fake or suspicious transactions.
Application
Patent
16. The computer program product of claim 15, wherein the discriminator network of the GAN comprises at least one of the following:
a first dense layer comprising a rectified linear unit (Relu) function with 128 nodes;
a second dense layer comprising a Relu function with 64 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 32 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 16 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a sigmoid function with 1 node, wherein the fifth dense layer is fully connected to the fourth dense layer;
a sixth dense layer comprising a sigmoid function with 1 node, wherein the sixth dense layer is fully connected to the fourth dense layer; or any combination thereof.
17. The computer program product of claim 16, wherein the discriminator network of the GAN comprises:
a first dense layer comprising a rectified linear unit (Relu) function with 128 nodes;
a second dense layer comprising a Relu function with 64 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a Relu function with 32 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 16 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a sigmoid function with 1 node, wherein the fifth dense layer is fully connected to the fourth dense layer; and
a sixth dense layer comprising a sigmoid function with 1 node, wherein the sixth dense layer is fully connected to the fourth dense layer.
Application
Patent
17. The computer program product of claim 16, wherein the one or more instructions further cause the at least one processor to:
during a training procedure, implement a dropout at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN; and
during the training procedure, implement a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN.
18. The computer program product of claim 17, wherein the one or more instructions further cause the at least one processor to:
during the training procedure, implement a dropout of 30% at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN; and
during the training procedure, implement a batch normalization process at each of the first dense layer, the second dense layer, the third dense layer, and the fourth dense layer of the discriminator network of the GAN.
Application
Patent
18. The computer program product of claim 16, wherein the one or more instructions further cause the at least one processor to:
provide an input to the discriminator network of the GAN and obtain an output that indicates whether an account is going to conduct a first payment transaction with a merchant of a plurality of merchants.
19. The computer program product of claim 16, wherein the one or more instructions further cause the at least one processor to:
provide an input to the discriminator network of the GAN and obtain an output that indicates whether an account is going to conduct a first payment transaction with a merchant of the plurality of merchants.
Application
Patent
19. The computer program product of claim 16, wherein one or more instructions further cause the at least one processor to:
optimize the discriminator network of the GAN based on a formula, wherein the formula includes a value of Recall for the discriminator network and the formula is defined as:
Recall= TP / (TP + FN),
wherein TP is a number of true positive predictions based on an output of the discriminator network corresponding to a ground truth label of a set of values of a plurality of features, and
wherein FN is a number of false negative predictions based on an output of the discriminator network not corresponding to a ground truth label of a set of values of a plurality of features.
20. The computer program product of claim 16, wherein one or more instructions further cause the at least one processor to:
during the training procedure,
optimize the discriminator network of the GAN based on a formula, wherein the formula includes a value of Recall for the discriminator network and the formula is defined as:
Recall= TP / (TP + FN);
wherein TP is a number of true positive predictions based on an output of the discriminator network corresponding to a ground truth label of a set of values of a plurality of features; and
wherein FN is a number of false negative predictions based on an output of the discriminator network not corresponding to a ground truth label of a set of values of a
plurality of features.
Application
Patent
20. The computer program product of claim 16, wherein the generator network of the GAN comprises at least one of the following:
a first dense layer comprising a rectified linear unit (Relu) function with 16 nodes;
a second dense layer comprising a Relu function with 32 nodes, wherein the second dense layer is fully connected to the first dense layer;
\a third dense layer comprising a Relu function with 64 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a Relu function with 128 nodes, wherein the fourth dense layer is fully connected to the third dense layer;
a fifth dense layer comprising a Relu function with 182 nodes, wherein the fifth dense layer is fully connected to the fourth dense layer; or
any combination thereof.
2. The computer-implemented method of claim 1, wherein the generator network of the GAN comprises:
a first dense layer comprising a rectified linear unit (ReLu) function with 16 nodes;
a second dense layer comprising a ReLu function with 32 nodes, wherein the second dense layer is fully connected to the first dense layer;
a third dense layer comprising a ReLu function with 64 nodes, wherein the third dense layer is fully connected to the second dense layer;
a fourth dense layer comprising a ReLu function with 128 nodes, wherein the fourth dense layer is fully connected to the third dense layer; and
a fifth dense layer comprising a ReLu function with 182 nodes, wherein the fifth dense layer is fully connected to the fourth dense layer.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. see PTO-892.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AHAMED I NAZAR whose telephone number is (571)270-3174. The examiner can normally be reached 10 am to 7 pm Mon-Fri.
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, Stephen Hong can be reached at 571-272-4124. 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.
/AHAMED I NAZAR/Examiner, Art Unit 2178 6/12/2026
/STEPHEN S HONG/Supervisory Patent Examiner, Art Unit 2178
Prosecution Insights