311 search results - page 4 / 63 » Gradient Convergence in Gradient methods with Errors |
COLT
2000
Springer
13 years 11 months ago
2000
Springer
We model reinforcement learning as the problem of learning to control a Partially Observable Markov Decision Process ( ¢¡¤£¦¥§ ), and focus on gradient ascent approache...
CDC
2009
IEEE
14 years 6 days ago
2009
IEEE
— In this paper we present a gradient method to iteratively update local controllers of a distributed linear system driven by stochastic disturbances. The control objective is to...
CEC
2008
IEEE
14 years 1 months ago
2008
IEEE
—This paper reviews the different gradient-based schemes and the sources of gradient, their availability, precision and computational complexity, and explores the benefits of usi...
MOC
2000
13 years 7 months ago
2000
Abstract. This paper is concerned with the mixed formulation of the Boussinesq equations in two-dimensional domains and its numerical approximation. The paper deals first with exis...
SCIA
2009
Springer
14 years 2 months ago
2009
Springer
Abstract. Level set methods are a popular way to solve the image segmentation problem in computer image analysis. A contour is implicitly represented by the zero level of a signed ...