Practical overview of image classification with tensor-network quantum circuits.

Area of Science:

• Quantum Computing
• Machine Learning
• Computational Science

Background:

• Designing effective quantum circuits for machine learning (QML) is a significant challenge.
• Tensor networks have shown promise in classical machine learning and offer a novel architectural basis for quantum circuits.

Purpose of the Study:

• To provide a comprehensive description of tensor-network quantum circuits (TNQCs) and their simulation.
• To demonstrate the application of TNQCs in image processing tasks for machine learning.

Main Methods:

• Detailed description and simulation methodology for tensor-network quantum circuits.
• Implementation using PennyLane, an open-source quantum differentiable programming library.
• Utilization of circuit cutting to enable simulation of larger quantum circuits on limited hardware.

Main Results:

• Successful simulation of various tensor-network quantum circuits.
• Demonstration of computational requirements and potential applications.
• Application of TNQCs to increasingly complex image processing tasks.

Conclusions:

• Tensor-network quantum circuits provide a flexible and powerful framework for QML.
• The described methods, including circuit cutting, facilitate the practical implementation and scaling of TNQCs.
• TNQCs show potential for addressing industrially relevant machine learning problems, particularly in image processing.