Contractual Quantum Deep Q-Learning

Increasingly, deep learning algorithms are leveraging the capabilities of quantum computing in a more complex form through quantum machine learning. This new approach significantly improves the processing and analysis of large data sets, optimizing complex algorithms with efficiencies unattainable by classical computing. Though, computation in the true sense of the word is often subject to constraints. This paper proposes Contract-based Quantum Deep Q-Learning that combines the rigor of constraint satisfaction with the expressive power of quantum learning. This is a crucial step in making Quantum Deep Q-Learning agents trustworthy and deployable in real systems. For a practical understanding of our approach and model, we carried out a simulation of a medical application related to the optimization of cancer treatment.