Mountaincar-v0-NN-Pytorch
This is a project for CMSC818B (Behavioral Planning in Robotics). A very famous problem of mountain carv0 solved using Neural networks.
- Cart try to reach at top but fails
- After training improvement in the learning where cart reach to the goal position
-
A policy which shows to push only to the one side(bad policy) and after training the updation of policy
-
bad policy
- after leraning process policy improves
- reward function shows that around 1000 iteration cart tries to reach at the top successfully.