406 search results - page 24 / 82 » Gadgets, Approximation, and Linear Programming |
IFIP
2005
Springer
14 years 3 months ago
2005
Springer
A new quasi-Newton scheme for updating a low rank positive semi-definite Hessian approximation is described, primarily for use in sequential quadratic programming methods for non...
POPL
2012
ACM
12 years 5 months ago
2012
ACM
Differential privacy is a notion of confidentiality that protects the privacy of individuals while allowing useful computations on their private data. Deriving differential priva...
ICTAI
2009
IEEE
14 years 4 months ago
2009
IEEE
Evolutionary computation methods have been used to solve several optimization and learning problems. This paper describes an application of evolutionary computation methods to con...
AAAI
2006
13 years 11 months ago
2006
Partially observable Markov decision processes (POMDPs) are an intuitive and general way to model sequential decision making problems under uncertainty. Unfortunately, even approx...
CDC
2010
IEEE
13 years 4 months ago
2010
IEEE
We consider the classical finite-state discounted Markovian decision problem, and we introduce a new policy iteration-like algorithm for finding the optimal state costs or Q-facto...