Description
- This course will study the algorithmic aspects of intelligent robotics and more generally autonomous systems.
- Core topics: planning, state estimation and learning that are central to an intelligent agent. Additionally, we will examine how the agent can make intelligent decisions given imperfect or uncertain knowledge about the world.
- Note that we will focus on the core algorithmic and AI aspects. The detailed physical interaction and control will largely be abstracted away for algorithmic treatment.
- Pre-requisites: Introduction to Artificial Intelligence (COL333-671) or Introduction to Machine Learning (COL774 or equivalent). Programming proficiency and knowledge of probabilistic models, deep learning, basic search algorithms, logic and probability will be an advantage.
- The course will have both practical and theoretical components. The practical component will build on theory and will require thorough experimentation and analysis by the students.
- This course is being offered the first time. It builds on the COL864 (Special Topics in AI) course offered at 8xx level. Students who have taken COL864 with the instructor should not register for COL778.
Logistics
| Class Days |
Mondays and Thursdays |
| Slot |
AA |
| Timing |
2:00 PM - 3:20 PM |
| Lecture Room |
LH 421 |
Announcements
- The remajor examination will be conducted on Wednesday, July 23rd 2025 (11am till 1pm in Bharti 501). This is only for students approved by the institute to take the examination. The instructions are here.
- The course is closed. Thanks for participating!
Topics
The class material will be covered on the white/black board. Slides and reference papers for some of the topics will also be provided. The list is tentative and can be slightly modified as we progress.
The class slides will be uploaded at this link.
| S. No. |
Topics |
| 1. |
Course Organisation |
| 2. |
Course Introduction |
| 3. |
Agent Representation - I |
| 4. |
Agent Representation - II |
| 5. |
State Estimation - I |
| 6. |
State Estimation - II |
| 7. |
Markov Decision Processes |
| 8. |
Task Planning |
| 9. |
Motion Planning |
| 10. |
RL Introduction |
| 11. |
Deep Q-Learning |
| 12. |
Imitation Learning |
| 13. |
Policy Gradients |
| 14. |
Actor Critic Methods |
| 15. |
Monte-Carlo Tree Search |
| 16. |
Partially-Observable MDPs |
References
- The primary reference for the course is the material covered in class.
- [AIMA] Artificial intelligence: a modern approach. Russell, Stuart J., and Peter Norvig.