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...
Importance sampling-based algorithms are a popular alternative when Bayesian network models are too large or too complex for exact algorithms. However, importance sampling is sensi...
Heterogeneous wireless sensor network (heterogeneous WSN) consists of sensor nodes with different ability, such as different computing power and sensing range. Compared with homoge...
Distributed embedded sensor networks are now being successfully deployed in environmental monitoring of natural phenomena as well as for applications in commerce and physical secu...
Richard Pon, Maxim A. Batalin, Mohammad H. Rahimi,...
This paper presents a system for recognizing sketched logic circuits in real-time and graphically simulating them afterwords. It has been developed for use in university and school...