Watch Priscila M.V. Lima discuss Artificial Intelligence meets Adiabatic Quantum Computing.
From the abstract:
We currently live in the era of machine learning wonders, to the extent that it is considered as a synonym for artificial intelligence by most lay people. However, some mishaps with classification and decision-making mechanisms have fomented an increasing demand for explanations to the output of those algorithms. One path to providing insight on machine learning results is that of generating interpretations based on analysis of features values. Another consists of embedding reasoning into machine learning in order to guarantee correct reasoning from learnt examples. A third, less explored alternative would be that of decoupling reasoning from learning from examples whilst maintaining the same neural environment. By associating optimal solutions to logical reasoning and other problems with energy minima, we introduce a solution to the specification of hard combinatorial problems as pseudo-Boolean constraints, SATyrus, which can also generate input to Adiabatic Quantum Computers, such as those that belong to the D Wave Computers Series.