Quantum-computer-based machine learning

Patent metadata

Field	Value
Publication number	US-12488273-B2
Application number	US-202117561804-A
Country	US
Kind code	B2
Filing date	Dec 24, 2021
Priority date	Dec 24, 2021
Publication date	Dec 2, 2025
Grant date	Dec 2, 2025

Abstract

Quantum computers with a limited number of input qubits are used to perform machine learning processes having a far greater number of trainable features. A list of features of a field are divided into a plurality of feature groups. Each of the feature groups includes a respective group of some, but not all, of the features. A first machine learning process is performed to train a first instance of a quantum computer model, where the feature groups are used as inputs. Based on the first machine learning process being performed, a subset of the feature groups is selected for a second machine learning process. Thereafter, the second machine learning process is performed to train one or more second instances of the quantum computer model. The individual features of the selected subset of the feature groups are used as inputs for the second instances of the quantum computer model.

First claim

What is claimed is:

1 . A method, comprising:
dividing a list of features of a field into a plurality of feature groups, such that each of the plurality of feature groups includes a respective group of some, but not all, of the features; performing a first machine learning process to train a first instance of a quantum computer model, the quantum computer model comprising a parameterized quantum circuit that includes a plurality of trainable single qubit gates and a plurality of fixed controlled-not (CNOT) gates, wherein the plurality of feature groups are used as inputs for the first instance of the quantum computer model in the first machine learning process such that a total number of the plurality of trainable single qubit gates of the first instance of the quantum computer model is less than or equal to a total number of the plurality of feature groups, and wherein the first machine learning process is performed at least in part by inputting each feature group of the plurality of feature groups into a respective one of the trainable single qubit gates of the first instance of the quantum computer model, wherein the parameterized quantum circuit further includes a measurement circuit that is configured to measure a result of a training of the trainable single qubit gates at least in part by measuring an electromagnetic frequency; based on the performing of the first machine learning process, selecting a subset of the plurality of feature groups for a second machine learning process; performing the second machine learning process to train one or more second instances of the quantum computer model, wherein individual features of the selected subset of the plurality of feature groups are used as inputs for the one or more second instances of the quantum computer model; based on the performing of the second machine learning process, selecting a subset of the features for additional machine learning training while discarding a rest of the features that are unselected; generating an updated list of features based on feature groups of the plurality of feature groups not selected based on performing the first machine learning process and the subset of the features selected based on performing the second machine learning process; and iteratively repeating the dividing, the performing the first machine learning process, the selecting the subset of the plurality of feature groups, the performing the second machine learning process, the selecting the subset of the features, and the generating the updated list of features, wherein the iteratively repeating is performed until a predefined number of features are selected based on performing the second machine learning process.

2 . The method of claim 1 , wherein the dividing comprises dividing the list of features based on: a principal component analysis, a random grouping, or a sequential grouping.

3 . The method of claim 1 , wherein the list of features includes features associated with a plurality of transactions.

4 . The method of claim 1 , further comprising: predicting fraud at least in part based on the performing of the second machine learning process.

5 . The method of claim 1 , wherein the first machine learning process or the second machine learning process is performed at least in part via a quantum neural network.

6 . The method of claim 1 , wherein the quantum computer model further comprises: an optimization circuit coupled to the parameterized quantum circuit, wherein the optimization circuit is configured to train the quantum computer model based on measured outputs of the parameterized quantum circuit.

7 . The method of claim 6 , wherein the optimization circuit is further configured to train the quantum computer model at least in part by minimizing a loss function defined as a part of the quantum computer model.

8 . The method of claim 1 , further comprising:
combining the discarded features from the second machine learning process into one or more new feature groups; performing a third machine learning process to train a third instance of the quantum computer model, wherein feature groups not selected by the first machine learning process and the one or more new feature groups are collectively used as inputs for the third instance of the quantum computer model in the third machine learning process; based on the performing of the third machine learning process, selecting a further subset of the plurality of feature groups for additional machine learning training; generating an updated list of features based on the further subset of the plurality of feature groups selected based on performing the third machine learning process and the subset of the features selected based on performing the second machine learning process; and iteratively repeating the dividing, the performing the first machine learning process, the selecting the subset of the plurality of feature groups, the performing the second machine learning process, the selecting the subset of the features, the combining, the selecting the further subset of the plurality of feature groups, and the generating the updated list of features, wherein the iteratively repeating is performed until a predefined number of features are selected based on performing the second machine learning process.

9 . A machine learning system, comprising: a first instance of a quantum computer machine learning model, wherein the first instance of the quantum computer machine learning model is configured to:
receive a plurality of feature groups as inputs, the plurality of feature groups each including a different plurality of individual features, respectively; perform a first machine learning process on the plurality of feature groups; and output, based on the first machine learning process, a subset of the plurality of feature groups for additional machine learning; one or more second instances of the quantum computer machine learning model coupled to the first instance of the quantum computer machine learning model, wherein the one or more second instances of the quantum computer machine learning model is each configured to: receive, as inputs, the plurality of individual features of a respective one of the plurality of feature groups from the subset of the plurality of feature groups outputted by the first instance of the quantum computer machine learning model; perform a second machine learning process on the plurality of individual features; and output, based on the second machine learning process, a subset of the individual features and a discarded subset of the plurality of feature groups; and a feature grouping module coupled to inputs of the first instance of the quantum computer machine learning model and coupled to outputs of the one or more second instances of the quantum computer machine learning model, wherein the feature grouping module is configured to: divide an initial list of features into the plurality of feature groups to be received by the first instance of the quantum computer machine learning model; receive the discarded subset of the plurality of feature groups as well as the subset of individual features outputted by the one or more second instances of the quantum computer machine learning model; re-divide the received discarded subset of the plurality of feature groups and the received subset of individual features into an updated list of feature groups; and output the updated list of feature groups to the inputs of the first instance of the quantum computer machine learning model; wherein the first instance and the one or more second instances of the quantum computer machine learning model each comprise a parameterized quantum circuit that includes a layer of trainable single qubit gate

Assignees

• Paypal Inc

Inventors

• Le Van Gong Hubert Andre
• Kumar Niraj
• Sharma Nitin S

Classifications

• G06N3/045

  Combinations of networks · CPC title

• G06F18/211

  Selection of the most significant subset of features · CPC title

• G06Q20/4016

  involving fraud or risk level assessment in transaction processing · CPC title

• G06N10/40

  Physical realisations or architectures of quantum processors or components for manipulating qubits, e.g. qubit coupling or qubit control · CPC title

• G06N3/08

  Learning methods · CPC title