Chapter 10
arash haratian
4/15/2021
Mountain Car task
source("./mountain_car.R")Figure 10.1: The Mountain Car task and the cost-to-go function
Interactive 3D plot
plot_fig10.1(plot_heatmap = FALSE)Heatmap plot
plot_fig10.1(plot_heatmap = TRUE)
Figure 10.2: Mountain Car learning curves for the semi-gradient Sarsa method with tile-coding
plot_fig10.2()
Figure 10.3: Performance of one-step vs 8-step semi-gradient Sarsa
plot_fig10.3()
Figure 10.4: Effect of the alpha and n on early performance of n-step semi-gradient Sarsa
plot_fig10.4()
An Access-Control queuing task
source("./queuing.R")Figure 10.5: The policy and value function found by differential semi-gradient one-step Sarsa
plot_fig10.5()