Abstract— A combination of backpropagation and neuroevolution is used to train a neural network visual controller for agents in the Quake II environment. The agents must learn to...
This paper discusses the empirical evaluation of improving generalization performance of neural networks by systematic treatment of training and test failures. As a result of syst...
— Solvers for the Boolean satisfiability problem are an important base technology for many applications. The most efficient SAT solvers for industrial applications are based on...
Cancers in head and neck tends to spread to nearby lymph nodes. Lymph nodes trap the spreading tumor cells but then the tumor starts to grow in these nodes and then spread further...
This study proposes a novel forecasting approach – an adaptive smoothing neural network (ASNN) – to predict foreign exchange rates. In this new model, adaptive smoothing techni...