Description: Off-Policy sampling method designed EM strategy Search for processing the continuous state and action high-dimensional control problems, for example, is a small ball balance problems.
To Search:
File list (Check if you may need any files):
EM strategy Search\basis.m
..................\collect_sample.asv
..................\collect_sample.m
..................\drawAction.asv
..................\drawAction.m
..................\drawnextstate.asv
..................\drawnextstate.m
..................\drawStartState.asv
..................\drawStartState.m
..................\ESREMPS.asv
..................\ESREMPS.m
..................\evalpol.m
..................\exp1_algorithm1.mat
..................\exp1_algorithm2(T=100).mat
..................\exp1_result(step100).mat
..................\exp2_algorithm1.mat
..................\exp2_algorithm2.mat
..................\exp2_result(step20).mat
..................\exp3_algorithm1.mat
..................\exp3_algorithm2.mat
..................\exp3_result(step50).mat
..................\fig1_reward(step100).fig
..................\fig2_reward( step20).fig
..................\fig3_reward(step50).fig
..................\getreward.m
..................\initPolicy.asv
..................\initPolicy.m
..................\IWPEEM.asv
..................\IWPEEM.m
..................\justexec.m
..................\LQR_learn.asv
..................\LQR_learn.m
..................\main.asv
..................\main.m
..................\plot_exp.asv
..................\plot_exp.m
..................\test1_alg1.mat
..................\test1_alg2.mat
..................\test2_alg1.mat
..................\test2_alg2.mat
..................\test3_alg1.mat
..................\test3_alg2.mat
..................\test_eval.asv
..................\test_eval.m
EM strategy Search
Download