# US 2020/0394067 A1 · Systems and methods for pre-executing idiosyncratic computation through the application of algorithmic prediction of user behavior patterns

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0394067 A1
- Document ID: US20200394067A1
- Status: Published application
- Kind: A1
- Application number: 16/442,796
- Filed: 2019-06-17
- Published: 2020-12-17
- Earliest claimed: 2019-06-17
- Assignees: Bank of America Corporation
- Applicants: Bank of America Corporation
- Inventors: Andrew Hosford; Bricklin Welch; Mohammad Athar
- Web page: https://patentgrep.com/patent/US20200394067A1

## CPC

- G06F 9/466: {Transaction processing}
- G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)}
- G06F 9/3802: {Instruction prefetching}
- G06F 16/219: {Managing data history or versioning (querying versioned data G06F16/2474; querying temporal data G06F16/2477)}
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/047: Probabilistic or stochastic networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn
- G06N 20/00: Machine learning
- G06Q 20/10: specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
- G06Q 20/40145: {Biometric identity checks}
- G07F 19/206: {Software aspects at ATMs}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G07F: COIN-FREED OR LIKE APPARATUS (coin sorting G07D3/00; coin testing G07D5/00; {handling coins or paper currencies apart from payment activated apparatus G07D; payment architectures, schemes or protocols G06Q20/00})

## Abstract

Aspects of the disclosure relate to a machine-learning transaction-prediction engine for seasoning an anticipated manual transaction. The transaction may occur in a transaction session. The seasoning may occur prior to execution of the anticipated manual transaction, The transaction-prediction engine may include a receiving/storage module configured to receive and store identification information of the transactor. The engine may also include a step-retrieval module configured to retrieve a set of in-session transaction steps associated with the transactor. In addition, the engine may include a history-retrieval module configured to retrieve, based on the stored identification information, historical transactional information associated with the transactor. The historical transaction information comprising a plurality of historical transaction patterns associated with the transactor. The engine may also include a processor module configured to initiate the anticipated manual transaction by predicting and provisionally completing the in-session transaction steps. The predicting may be based on the set of in-session transaction associated with the transactor and on the plurality of historical patterns.

## Claims

### Claim 1

A computational logic for pre-fetching of information related to predicting a response to an anticipated manual transaction of a transactor, the transaction occurring in a transaction session, said pre-fetching occurring prior to execution of the anticipated manual transaction, said computational logic comprising:
receiving and storing identification information of the transactor;
retrieving a set of in-session transaction steps associated with the transactor;
based on the stored identification information, retrieving historical transactional information associated with the transactor, said historical transaction information comprising a plurality of historical transaction patterns associated with the transactor; and
based on the set of in-session transaction steps associated with the transactor and on the plurality of historical patterns, initiating the anticipated manual transaction by predicting and provisionally completing the in-session transaction steps.

### Claim 2 (depends on 1)

The computational logic of claim 1, wherein the receiving and storing are triggered by the transactor commencing the manual transaction.

### Claim 3 (depends on 2)

The computational logic of claim 2, wherein the commencing the manual transaction includes entering a code into a transaction system, the code identifying the transactor.

### Claim 4 (depends on 1)

The computational logic of claim 1, wherein the receiving and storing are triggered by the matching, with a pre-determined confidence level, of received identifying information of the transactor with a saved identification profile associated with the transactor.

### Claim 5 (depends on 4)

The computational logic of claim 4, wherein the received identifying information of the transactor is at least augmented by biometric information provided via one or more than one biometric device.

### Claim 6 (depends on 4)

The computational logic of claim 4, wherein the historical transaction patterns associated with the transactor, in conjunction with a time stamp of the transactor commencing the manual transaction, are processed by the computational logic to predict the transaction type associated with the transaction.

### Claim 7 (depends on 1)

The computational logic of claim 1, wherein the set of in-session transaction steps includes selection by the transactor of a transaction type associated with the manual transaction, prior to the transactor specifying a region associated with the manual transaction.

### Claim 8 (depends on 1)

The computational logic of claim 1, wherein the set of in-session transaction steps includes selection by the transactor of a transaction type associated with the manual transaction, prior to the transactor specifying a range of dates associated with the manual transaction.

### Claim 9 (depends on 1)

The computational logic of claim 1, wherein the set of in-session transaction steps includes selection by the transactor of a transaction type associated with the manual transaction, prior to the transactor specifying an entity associated with the manual transaction.

### Claim 10 (depends on 1)

The computational logic of claim 1, wherein:
the set of in-session transaction steps includes a selection by the transactor of a not yet completed sequence of information, the full sequence to specify a region and range of dates associated with the transaction; and
the historical transaction patterns associated with the transactor, in conjunction with the not yet completed selection of the sequence of numerically input digits are processed by the computational logic to predict the full sequence of digits specifying the amount associated with the transaction.

### Claim 11 (depends on 1)

The computational logic of claim 1, further comprising:
presenting the transactor with an output of the provisionally completed in-session transaction steps;
in response to confirmation by the transactor of correctness of the output provisionally completed in-session transaction steps as accurately predicting the anticipated manual transaction, executing the transaction as provisionally completed; and
otherwise, cancelling the transaction as provisionally completed, awaiting transactor input of further manual input of in-session transaction steps, and executing the transaction as manually input by the transactor.

### Claim 12 (depends on 10)

The computational logic of claim 10, further comprising:
augmenting the historical transactional information associated with the transactor with data of the executed transaction;
saving a measure of the confirmation or otherwise by the transactor of the correctness of the provisionally completed in-session transaction steps as accurately predicting, via machine learning, the anticipated manual transaction; and
utilizing a neural network to use the saved measure to further train the computational logic toward improving the predictive power of the computational logic.

### Claim 13 (depends on 12)

The computational logic of claim 12 wherein the measure of the confirmation or otherwise of the correctness of the provisionally completed transaction, reflects at least one of:
a point along the transaction at which the computational logic processed the confirmed prediction; and
a degree of mismatch between the prediction and the transaction as manually input by the transactor.

### Claim 14

A machine-learning transaction-prediction engine for pre-fetching of information related to predicting a response to an anticipated manual transaction, the transaction occurring in a transaction session, said pre-fetching occurring prior to execution of the anticipated manual transaction, said transaction-prediction engine comprising:
a receiving/storage module configured to receive and store identification information of the transactor;
a step-retrieval module configured to retrieve a set of in-session transaction steps associated with the transactor;
a history-retrieval module configured to retrieve, based on the stored identification information, historical transactional information associated with the transactor, said historical transaction information comprising a plurality of historical transaction patterns associated with the transactor; and
a processor module configured to initiate the anticipated manual transaction by predicting and provisionally completing the in-session transaction steps, said predicting based on the set of in-session transaction steps associated with the transactor and on the plurality of historical patterns.

### Claim 15 (depends on 14)

The transaction-prediction engine of claim 14, wherein the receiving/storage module is triggered to receive and store the identification information by the matching, with a pre-determined confidence level, of received identifying information of the transactor with a saved identification profile associated with the transactor.

### Claim 16 (depends on 15)

The transaction-prediction engine of claim 15, wherein the processor module is configured to predict the transaction type associated with the transaction based on the historical transaction patterns associated with the transactor in conjunction with a time stamp of the transactor commencing the manual transaction.

### Claim 17 (depends on 14)

The transaction-prediction engine of claim 14, further comprising:
an input/output module configured to:
present the transactor with an output of the provisionally completed in-session transaction; and
receive confirmation or otherwise from the transactor of correctness of the output provisionally completed in-session transaction as accurately predicting the anticipated manual transaction; and
an execution/cancellation module configured to:
in response to confirmation by the transactor of correctness of the output provisionally completed in-session transaction, execute the transaction as provisionally completed; and
otherwise, cancel the transaction as provisionally completed and execute the transaction as manually input by the transactor,

### Claim 18 (depends on 17)

The transaction-prediction engine of claim 17, further comprising:
an augmentation module configured to augment the historical transactional information associated with the transactor with data of the executed transaction;
a session-assessment module configured to store a measure of the confirmation by the transactor of the correctness of the provisionally completed in-session transaction; and
a neural network configured to use the stored measure to train the transaction-prediction engine toward improving the predictive accuracy of the transaction-prediction engine.
