Optimal Control, Reinforcement Learning & Robotics Projects in CS 287: Advanced Robotics (fall 2019)

This repository contains past projects I've completed in CS 287, which I took in Berkeley EECS department under Prof. Pieter Abbeel, in fall 2019. (https://people.eecs.berkeley.edu/~pabbeel/cs287-fa19/)

Followings are those projects and related materials covered/implemented in each project.

• Project 1

  • Value Iteration
  • Dicretization-based Optimal Control
  • Function Approximation Optimal Control

• Project 2

  • LQR, iLQR
  • DDP
  • Feedback Linearization

• Project 3

  • Convex Optimiazation
  • Sequential Convex Programming
  • Motion Planning and Control (w/ convex opt.)

• Project 4

  • Multivariate Gaussians
  • Kalman Filtering
  • EM & MLE
  • Particle Filtering
  • Belief Space Planning

• Project 5

  • Policy Gradient
    • Trust Region Policy Optimization (TRPO)
    • Proximal Policy Optimization (PPO)
  • Deep Q-Learning
    • DQN
    • Double DQN
    • Dueling DQN