Chris Pollett > Students > Garg

    (Print View)



    [CS 297 Proposal]

    [Q-Learning-Presentation - PDF]

    [Deliverable 1]

    [FABRIK-Presentation - PDF]

    [Deep-Q-Networks-Presentation - PDF]

    [Deliverable 2]

    [TRPO-Presentation - PDF]

    [Deliverable 3]

    [DDPG-Presentation - PDF]

    [Deliverable 4]

    [CS297 Report - PDF]

    [CS298 Proposal]

    [CS298 Report - PDF]

Project Blog

Meeting 10: April 17, 2018
Gave presentation on DDPG paper
Tasks assigned:
  • Add screenshots for previous deliverables.
  • Work on Deliverable 4.
Meeting 9: April 11, 2018
Presented Deliverable 3.
Tasks assigned:
  • Read DDPG paper
  • Work on Deliverable 4. Scope of Deliverable 4 can be limited to making the humanoid not fall down.
Meeting 8: April 03, 2018
Gave part 1 of presentation on "Trust region policy optimization" paper.
Tasks assigned:
  • Make the rock climbing pattern more realistic.
  • Complete "Trust region policy optimization" presentation
  • Work on Deliverable 3. Explore mujoco-py
Meeting 7: Mar 20, 2018
Presented Deliverable 2. Also gave presentation on "Playing Atari with Deep Reinforcement Learning" paper.
Tasks assigned:
  • Make the rock climbing pattern more realistic.
  • Read "Trust region policy optimization" paper
  • Work on Deliverable 3

Meeting 6: Mar 6, 2018
Gave presentation on FABRIK paper. Also showed a very simple demo of a box with a single rock-wall grip.
Tasks assigned:
  • Work on Deliverable 2. Script should be able to create random rock climbing structures.
  • FABRIK provided the final posture, but moving to that posture is a problem in itself(e.g. obstacles). Figure out how to do that.

Meeting 5: Feb 27, 2018
Gave presentation on Strange Beta paper. Decided that it was not very relevant to our project. Also discussed Atari paper and TD-Gammon paper.
Tasks assigned:
  • Read paper "FABRIK: A fast, iterative solver for the Inverse Kinematics problem"
  • Explore simulation tools and create a basic wall structure.

Meeting 4: Feb 20, 2018
Gave Deliverable 1: Q-Learning implementation for Tic-Tac-Toe game
Tasks assigned:
  • Read paper "Strange Beta: An Assistance System for Indoor Rock Climbing Route Setting Using Chaotic Variations and Machine Learning"
  • Read paper "Playing Atari with Deep Reinforcement Learning"
  • Read paper "Temporal DiWerence Learning and TD-Gammon"

Meeting 3: Feb 13, 2018
Gave presentation on Q-Learning and discussed the next deliverable.
Tasks assigned:
  • Deliverable 1: Q-Learning algorithm implementation for Tic Tac Toe game

Meeting 2: Feb 6, 2018
Discussed and finalised proposal details.
Tasks assigned:
  • Update Bio, Proposal and blog
  • Prepare presentation on Q-learning

Meeting 1: Jan 30, 2018
Organisational meeting. Finalised meeting time to be between 1:30 PM and 2:15 PM
Tasks assigned:
  • Prepare project proposal