6.891: Learning in Worlds with Objects
Most machine-learning techniques assume a representation of states of
the problem in terms of vectors of discrete or real-valued
attributes. We would like to apply machine learning to more complex
situations that humans would describe in terms of the objects that compose
them (chairs, tables, books, people). However, it is not
clear how to represent such domains in order to afford effective
generalization, nor what algorithms to use to learn these more complex
representations.
This course will be a seminar in which we study research papers from
related areas and try to develop one or more approaches to
representation and learning in worlds with objects.
- Time: Spring 2002: Thursdays 2 - 4
- Place: 34-304
- Instructor: Leslie Pack Kaelbling
- Prerequisites: A previous graduate-level course in AI;
familiarity with basic techniques from machine learning and logic.
- Course work: Students must read all the papers; additionally,
they will lead one paper discussion
and complete a final paper or project, with in-class presentation
- Listeners: Listeners are welcome, on the condition that they
attend most of the sessions and read the papers under discussion
Syllabus
Feb 7:
Varieties of learning and representation
Required:
- Sarah Finney, Natalia Hernandez Gardiol, Tim Oates, and Leslie
Pack Kaelbling, "Learning in
Worlds with Objects." Working Notes of
the AAAI Stanford Spring Symposium on Learning Grounded
Representations, 2001.
- Stuart J. Russell, "Execution architectures and compilation." In
Proceedings of the Eleventh International Joint Conference on
Artificial Intelligence, Detroit, MI: Morgan Kaufmann, 1989.
Optional:
- Leslie Pack Kaelbling, Michael L. Littman, and Andrew W. Moore,
"Reinforcement
Learning: A Survey." Journal of Artificial Intelligence
Research, Volume 4, 1996.
Feb 14:
Deterministic worlds with hidden state
Required:
- Dana Angluin. "Learning regular sets from queries and
counterexamples." Information and Computation, 75(2):87-106,
November 1987.
- Ronald L. Rivest and Robert E. Schapire.
"Inference of finite automata using homing sequences."
Information and Computation, 103(2):299-347, April
1993.
Optional:
- Ronald L. Rivest and Robert E. Schapire.
"Diversity-based inference of finite automata."
Journal of the Association for Computing Machinery,
41(3):555-589, May 1994.
Feb 21:
Stochastic worlds with hidden state
Required:
- Lawrence R. Rabiner.
"A Tutorial on Hidden Markov Models and Selected Applications in
Speech Recognition."
Proceedings of the IEEE, 77(2):257-286, February 1989.
- Michael Littman, Richard Sutton, and Satinder Singh.
"Predictive Representations of State."
Neural Information Processing Systems, 2001.
- Gary Drescher.
"A Mechanism for Early Piagetian Learning."
In Proceedings of the National Conference on Artificial
Intelligence, 1987.
Optional:
Feb 29 (Friday!!):
Deterministic relational representations
-
Ronald J. Brachman and Hector J. Levesque, Knowledge
Representation and Reasoning, draft, 2001, Chapters 1, 9, and 16.
March 7:
Inductive logic programming
Required:
Optional:
March 14:
Situated logical learning
Required:
-
Saso Dzeroski, Luc De Raedt, and Hendrik Blockeel, "Relational
Reinforcement Learning." Proceedings of the 8th
International Conference on Inductive Logic Programming, Lecture Notes
in Artificial Intelligence, Vol. 1446, Springer, 1998.
- Kurt Driessens, J
an Ramon, and Hendrik Blockeel, "Speeding
up Relational Reinforcement Learning Through the Use of an Incremental
First Order Decision Tree Learner." European Conference on
Machine Learning, 2001.
March 21:
Probabilistic relational representations
Required:
- Kristian Kersting and Luc de Raedt, "Bayesian Logic
Programs." In J. Cussens and A. Frisch, editors,
Work-in-Progress Reports of the Tenth International Conference on
Inductive Logic Programming, London, 2000.
Optional:
March 28:
Spring Break
April 4:
Probabilistic relational learning
-
Nir Friedman, Lise Getoor, Daphne Koller, and Avi Pfeffer,
"Learning
Probabilistic Relational Models." In
Proceedings of the Sixteenth International joint Conference on
Artificial Ingelligence, 1999.
-
Lise Getoor, Nir Friedman, Daphne Koller, and Ben Taskar,
"Learning
Probabilistic Models of Relational
Structure." In Proceedings of the International Conference on Machine Learning, 2001.
April 11: More probabilistic relational learning
Required:
Optional:
April 12 (Friday!!):
Logic and the real world: Symbol grounding
April 18: No class!
April 25: More grounding, for real
-
Philip E. Agre and David Chapman, "Pengi: An Implementation of a
Theory of Activity," in Proceedings of the Sixth National Conference
on Artificial Intelligence, 1987.
-
Ian D. Horswill,
"Grounding
Mundane Inference in Perception."
Autonomous Robots, Volume 5, 1998.
-
Silvia Coradeschi and Alessandro Saffiotti,
"Perceptual
Anchoring of Symbols for Action."
In Proceedings of the Seventeenth International Joint Conference on
Artificial Intelligence, 2001.
May 2: Signals to symbols: learning to interpret your sensors
May 9: Signals to symbols: object representation
-
Heinrich H. Buelthoff, Shimon Y. Edelman, and Michael J. Tarr,
How are three-dimensional objects represented in the
brain?, Artificial Intelligence Memo AIM1479, Massachusetts
Institute of Technology, April 1994.
- Shimon Ullman, High-level Vision, Chapters 1 and 2,
The MIT Press, 1996.
May 10: Project Presentations
- Brenda Ng
- Pearlin Cheung
- Nick Hanssens
- Tracy Hammond
- Yuan Qi
May 16:
Project presentations
- Sarah Finney
- Hesky Fisher
- Jake Beal
- Natalia Hernandez Gardiol
- Mike McGeachie
- Cory Kidd