MIT CSAIL6.882 Embodied Intelligence |
||
Spring 2019 |
||
This schedule is preliminary and subject to change as the term evolves.
| Date | Topics | Course Materials | Assignments | ||
| Week 1: Cognitive architectures and paradigms | |||||
| Tues 2/5/2019 | Domains, agent architecture, factory/wild | Russell & Norvig slides | |||
| Thurs 2/7/2019 | Classic papers |
Assigned readings: Shakey (video)(Technical Report) Brooks: Elephants Don't Play Chess Sims: Evolved Virtual Creatures (video) Sutton: Dyna |
Assignment #1 | ||
| Week 2: The problem of long-term decision making | |||||
| Tues 2/12/2019 | Deterministic domains: pddl planning, IW(k), power of heuristics, factoring |
Assigned readings: Bonet & Geffener: HSP Lipovetsky & Geffner: IW(k) Bandres, Bonet & Geffner: Planning with Pixels |
Assignment #2 | ||
| Thurs 2/14/2019 | Future-state uncertainty: value iteration, mcts, MPC / replanning |
Assigned readings: Kocsis & Szepesvari: Bandit-based Monte-Carlo Planning Yoon, Fern & Givan: FF-Replan: A Baseline for Probabilistic Planning Riedmiller: Neural Fitted Q Iteration |
Assignment #3 | ||
| Week 3: The problem of continuous action spaces | |||||
| Thurs 2/21/2019 | Motion planning, optimization |
Assigned readings: Kuffner & LaValle: RRT-Connect (animations) Garrett et al.: Factored Transition Systems (video and journal version) Mordatch et al.: Contact-Invariant Optimization (video and video) |
Assignment #4 | ||
| Week 4: The problem of situation awareness | |||||
| Tues 2/26/2019 | Perception as state estimation |
Assigned readings: MaskRCNN Learning to See Physics via Visual De-animation The feeling of success |
Assignment #5 | ||
| Thurs 2/28/2019 | Sensory representation learning |
Assigned readings: Large Scale Unsupervised Learning (cat face paper) Learning image representations tied to ego-motion Contrastive predictive coding |
Assignment #6 | ||
| Week 5: The problem of partial observability | |||||
| Tues 3/5/2019 | POMDPs: search and policies |
Assigned readings: POMDP Tutorial Despot: Online POMDP Planning Belief-space Planning |
Assignment #7 | ||
| Thurs 3/7/2019 | POMDPs: Learning |
Assigned readings: RL in Robotics: A Survey Solving Deep Memory POMDPs Deep Recurrent Q-Learning for POMDPs |
Assignment #8 | ||
| Week 6: The problem of learning for long horizons | |||||
| Tues 3/12/2019 | Exploration, novelty search / curiosity, intrinsic motivation |
Assigned readings: Novelty #Exploration Curiosity-Driven Exploration |
Assignment #9 | ||
| Thurs 3/14/2019 | HRL |
Assigned readings: Feudal RL h-DQN Data-Efficient HRL |
Assignment #10 | ||
| Week 7: The problem of very little data | |||||
| Tues 3/19/2019 | Model-based intelligence | Assigned readings: Building Machines That Learn and Think Like People (read Section 4, skim rest) Deep Visual Foresight World Models |
Assignment #11 | ||
| Thurs 3/21/2019 | Meta-learning | Assigned readings: MAML RL^2 Modular Meta-Learning |
Assignment #12 | ||
| Thurs 3/21/2019 | Project | Assigned readings: Project handout |
|||
| Week 8: The problem of nonstationarity | |||||
| Tues 4/2/2019 | Methodology | Assigned readings: Comparing Classifiers (section 1 only) Machine Learning that Matters Deep RL that Matters |
Assignment #13 | ||
| Thurs 4/4/2019 | Multiagent RL | Assigned readings: Comprehensive survey of MARL QMIX Learning to Communicate with Deep MARL |
Assignment #14 | ||
| Week 9: The problem of adversaries | |||||
| Tues 4/9/2019 | Adversaries/Self-play | Assigned readings: Markov games TD-Gammon AlphaZero |
Assignment #15 | ||
| Thurs 4/11/2019 | Artificial life | Assigned readings: Karl Sims part 2 (evolving by competition) Emergent complexity via multiagent competition Deep neuroevolution |
Assignment #16 | ||
| Week 10: Concrete problems I | |||||
| Thurs 4/18/2019 | A general-purpose robot hand | Assigned Reading: Learning Dexterous In-Hand Manipulation (Movie) |
|||
| Week 11: Concrete problems II | |||||
| Tues 4/23/2019 | Making tea at in anybody's house | Assigned Reading: Exploiting Probabilistic Knowledge under Uncertain Sensingr |
|||
| Thurs 4/25/2019 Meet in 56-154 |
Assembling Ikea furniture | Assigned Reading: Can robots assemble an IKEA chair? (Supplementary material, including movie) |
|||
| Week 12: Project milestone presentations | |||||
| Tues 4/30/2019 | Project milestone presentations | ||||
| Thurs 5/2/2019 | Project milestone presentations | ||||
| Week 13: Work on project, no class meetings | |||||
| Week 14: Final project presentations | |||||
| Tues 5/14/2019 | Final project presentations | ||||
| Thurs 5/16/2019 | Final project presentations | Final project due | |||