Quantum Circuit Training with Fine-Tuned Models for Data Classification

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing quantum circuits face challenges in improving classification accuracy due to low relevance between training datasets, increased training time, and limitations of current quantum computers, such as limited qubit count and high error frequencies.

Innovation Solution

A method involving a trained model that is fine-tuned using a specific dataset to improve classification accuracy, followed by training a quantum circuit based on the updated model to enhance feature extraction and classification results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a quantum circuit is trained directly using existing methods, then training can be performed, but classification accuracy is low due to low relevance between training datasets

Engineering Contradiction:

Improveclassification accuracyVSAvoidrelevance between datasets

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

A classical trained model is introduced as an intermediary between the dataset and the quantum circuit. The trained model processes the dataset and generates enhanced feature amounts that are then fed to the quantum circuit, improving classification accuracy by bridging the relevance gap between datasets

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The classification system is divided into two separate components: a classical trained model for feature extraction and a quantum circuit for final classification. This segmentation allows each component to specialize in its strengths, with the trained model handling data preprocessing and the quantum circuit handling quantum-enhanced classification

Inventive Principle:

Principle #1Segmentation

2Measurement precision

If training data is extensively processed to improve accuracy, then classification accuracy improves, but training time increases

Engineering Contradiction:

Improveclassification accuracyVSAvoidtraining time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

Feature extraction and data preprocessing are performed in advance by the trained model before quantum circuit training begins. This preliminary action prepares optimized input data that reduces the computational burden during quantum circuit training, thereby reducing overall training time while maintaining accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system maintains continuous useful action by having the trained model continuously process and transform data into optimized feature representations that are immediately available for quantum circuit training, eliminating idle time and ensuring efficient utilization of computational resources throughout the training process

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS20250307677A1Computer-readable recording medium storing information processing program, information processing method, and information processing apparatus

Publication Date: 2025.10.02 FUJITSU LTD

AI summary

A non-transitory computer-readable recording medium storing an information processing program causing a computer to execute processing including: acquiring a trained model that has a function of outputting a classification result of input data in accordance with a feature amount extracted from the data; updating the acquired trained model based on a dataset that includes specific type of data such that classification accuracy of the specific type of data is improved; and controlling an arithmetic unit to train, based on the dataset, a quantum circuit that has a function of outputting a classification result of the specific type of data in accordance with a feature amount extracted from the specific type of data by the updated trained model.