Constraint satisfaction consistency preprocessing methods are used to reduce search e ort. Time and especially space costs limit the amount of preprocessing that will be cost e ec...
It is widely accepted that the use of more compact representations than lookup tables is crucial to scaling reinforcement learning (RL) algorithms to real-world problems. Unfortun...
Satinder P. Singh, Tommi Jaakkola, Michael I. Jord...
models require the identi cation of abstractions and approximations that are well suited to the task at hand. In this paper we analyze the problem of automatically selecting adequ...
A homogeneous paradigm for evidence integration is presented, and a vision system to recognize 3D objects is demonstrated using this paradigm. A new concept called generalizedfeat...
Rick Kjeldsen, Ruud M. Bolle, Andrea Califano, Rus...
In the past several years, there have been a large number of published works that have graphically described a wide variety of security problems particular to Unix. Without fail, ...