Diagrammatic Differentiation for Quantum Machine Learning
Diagrams are becoming a prominent tool in both machine learning and quantum computing. We adapt a key tool of machine learning, gradients, to general diagrammatic theories. This will enable one to do (quantum) machine learning fully diagrammatically, substantially broadening the road towards general quantum advantage, and quantum natural language processing in particular.
Alexis Toumi, Richie Yeung, Giovanni de Felice