COL-864: Special Topics in Artificial Intelligence
Planning and Estimation for Autonomous Systems
Credits: (3-0-0)
Holi Term 2021
Description
Planning and estimation are central to autonomous systems operating in the real world. This course will cover the concepts, principles and methods for intelligent decision-making with imperfect or uncertain knowledge. Students will develop an understanding of how different planning and learning techniques are usefulin problem domains where robots or other embodied-AI agents are deployed. Introduction to Artificial Intelligence (COL333-671) or Introduction to Machine Learning (COL774 or equivalent). Programming proficiency and knowledge of probabilistic models, basic deep learning, basic search algorithms, logic and probability will be an advantage.
Announcements
- Welcome to the course. First class on February 4, 2019 over MS Teams.
- Class videos are uploaded on Impartus.
- Assignment 1 is released on March 02, 2021. Submission is due on March 21, 2021.
- Minor exam is scheduled for March 19, 2021. Exam logistics and submission information appears below.
- The submission deadline for Assignment I is revised to 5pm on March 24, 2021. No late submissions.
- Assignment 2 is released on April 17, 2021. Submission is due on May 02, 2021.
- The last lecture for the course was delivered on April 27, 2021. The lecture material is now complete.
- As per the directions of Dean (Academics) notified on April 29, 2021 the pass criterion is uniformly set to 30% aggregate marks for both audit and credit.
- The submission deadline for Assignment II is extended for all students till 5pm on May 05, 2021. Please see the note below.
- Major exam is scheduled for May 10, 2021. Exam logistics and submission information appears below.
Course Information
- Instructor: Rohan Paul
- Classes: Slot AD
- Teaching Assistant: Vikas Upadhyay (vikas.upadhyay@cse.iitd.ac.in)
Lectures
S. No. | Topic | Class Material |
---|---|---|
1 | Course Organization | Slides |
2 | Course Introduction | Slides |
3 | Agent Representation | Slides |
4 | Planning Motions | Slides |
5 | State Estimation - I | Slides |
6 | State Estimation - II | Slides |
7 | Planning - A* Search | Slides |
8 | Task Planning | Slides |
9 | Markov Decision Processes | Slides |
10 | Model-Based RL | Slides |
11 | Model-Free RL - II | Slides | 12 | DQN and Policy Gradients | Slides | 13 | Partially-Obervable MDPs | Slides |
Revised Pass Criterion
- 30% aggregate marks for students taking the course for credit (D grade)
- 30% aggregate marks for students taking the course for audit (NP grade)
- The above is as per directions of Dean Academics notified on April 29, 2021.
Revised Assignment II Submission
- The submission deadline for Assignment II is extended for all students till 5pm on May 05, 2021.
- The ongoing pandemic situation has affected students in different ways. Several students facing infection have already contacted the instructor and the TA.
- The instructor will assist (given constraints) to enable recovering students to make their submissions. In case a student recovering from the disease requires additional time period beyond May 05, 2021 to complete his/her submission, the student is requested to contact the instructor directly over email. Assistance will be provided on a case to case basis.
Assignments
- Assignment I. Submission due at 5pm on March 24, 2021. No late submissions please.
- Assignment II. Submission due at 5pm on May 05, 2021. No late submissions please.
Examination
- Minor examination is scheduled for Friday March 19, 2021. Review guidelines at the following link.
- Major examination is scheduled for Monday May 10, 2021. Review guidelines at the following link.
References
- [AIMA] Artificial intelligence: a modern approach. Russell, Stuart J., and Peter Norvig. Link.
- [PR] Probabilistic robotics. Thrun, Sebastian, Wolfram Burgard, and Dieter Fox. Link. Online.
- [DM] Mykel Kochenderfer,Decision Making Under Uncertainty
- [DL] Deep Learning. Ian Goodfellow, Yoshua Bengio and Aaron Courville. Online [DL]
- [PA] Planning Algorithms. LaValle, S.M., 2006. Cambridge University Press. Online.
- [SB] Reinforcement Learning (Second Edition). Richard Sutton and Andrew Barto. MIT Press. 2018. Online.
Background Reading Material
- Pointers for reviewing some of the background topics for the course. Some of the material may be briefly in class.
- Classical Planning (AIMA Ch. 3)
- Neural Networks (DL Ch. 6)
- Markov Decision Processes (AIMA Ch 17.1-17.3 )
- Probabilistic Models (AIMA Ch 14.1-14.5)
- These are starting pointers but not an exhaustive list, you are welcome to explore further.
Learning outcomes
At the end of the course students will be able to: model autonomous systems as AI agents, formulate/solve relevant planning/estimation tasks. Further, students will gain insights in the computational challenges arising from uncertainty and how to incorporate recent learning-based methods decision-making algorithms.