Reinforcement learning (RL) problems constitute an important class of learning and control problems faced by artificial intelligence systems. In these problems, one is faced with ...
Abstract. We present a generalization of the Perceptron algorithm. The new algorithm performs a Perceptron-style update whenever the margin of an example is smaller than a predefi...
We describe an extension to the Mixture of Experts architecture for modelling and controlling dynamical systems which exhibit multiple modesof behavior. This extension is based on...
One of the most challenging aspects of managing a very large data warehouse is identifying how queries will behave before they start executing. Yet knowing their performance charac...
Archana Ganapathi, Harumi A. Kuno, Umeshwar Dayal,...
: I will discuss two efforts to get Intrusion Detection Systems to work together - the Common Intrusion Detection Framework (CIDF), and the IETF's working group to develop an ...