MEASURE CHANGE TECHNIQUES IN OPTIMAL-CONTROL

Fully and partially observed stochastic control of systems with non-dy namics and terminal and running costs are considered. Measure changes are introduced which allow both state and observation dynamics to be t hought of as linear. In the case when the terms of the cost have a spe cial form the measure change transformation ''cancels out'' the nonlin earities and changes the original nonlinear problem into a classical L QG one and standard results can be applied, We also consider unnormali zed conditional densities of the whole path as state variables and obt ain dynamic programming and verification results.