Image Classification Using Quantum Machine Learning

Abstract

When quantum algorithms are used in machine learning systems, it is referred to as “quantum machine learning.” An approach known as “quantum-enhanced machine learning” utilizes a quantum computer to evaluate classical data to boost machine learning. Data may be processed and stored more quickly and efficiently with the help of quantum machine learning. Using neural networks as analogies for physical systems is an important part of quantum machine learning. This paper summarizes the CIFAR-10 dataset. For the dataset, “five training batches and one testing batch” are used to divide the ten thousand photographs. One thousand images from each class are randomly selected for inclusion in the test batch. Even though each batch comprises all of the remaining photographs, some batches have a greater number of images from a particular category. It is estimated that each training batch contains around 5000 photographs. This section includes an evaluation of the classifier’s overall performance. Quantum neural networks describe “a parameterized quantum computational model best” implemented “on a quantum computer” (QNN). Third-party libraries such as PyTorch, Qiskit, and matplotlib are frequently loaded into the program. PyTorch is a popular option for GPU and CPU-based Deep Learning applications because it is built on Torch rather than merely Python. All of your quantum computing needs may be met by Qiskit, a Python library. Importing it will be necessary after the system is installed. Creating static, animated, and interactive graphics is easy with Matplotlib, a Python toolkit. To begin, we need to identify the quantum layers that will make up the circuit’s structure.