The world’s first neural network able to explain itself is in development at Raytheon as part of the DARPA Explainable Artificial Intelligence programme. Through the programme DARPA aims to establish a suite of machine learning techniques with the ability to produce explainable models without compromising on performance. This will allow human ...