AutoQML: Automatic Generation and Training of Robust Quantum-Inspired Classifiers by Using Genetic Algorithms on Grayscale Images

Some of the computation paradigm is based on the ability to use physical proposed advantages of QML include access to larger feature properties such as entanglement or superposition, which allow spaces, more general and expressive models and algorithmic quantum bits or qubits, which are basic information units in improvements in model optimization. Our proposed quantum quantum computing, to be in more than one state at the same image classification methods utilize these advantages. We time (see Section II-A), allowing access to Hilbert spaces and optimize the models by using metaheuristic techniques to thus to spaces that may be infinite-dimensional H. Similar to automatically obtain the best predictions based on test data, classical computing, in which information is calculated based ensuring model robustness, and we automatically generate on electrical circuits and logic gates that operate on bits, in simple quantum-inspired machine learning classifiers that can quantum computing, quantum circuits composed of sequences easily be implemented on classical computers. of quantum gates are used to modify quantum states (see In addition to algorithms that are useful for future scalable Section II-B).