Outline • Course wrap up • Final exam info Course wrap up • Other AI courses • AI jobs July 25, 2006 • AI research CS 486/686 University of Waterloo 2 CS486/686 Lecture Slides (c) 2006 P. Poupart Rational Agents Agents and Environments • Recall: A rational agent “does the right thing” • Performance measure – success criteria sensors – Evaluates a sequence of environment states percepts environment ? • A rational agent chooses whichever action agent maximizes the expected value of its actions performance measure given the percept actuators sequence to date – Need to know performance measure, environment, Agents include humans, robots, softbots, thermostats… possible actions, percept sequence The agent function maps percepts to actions f:P* � A The agent program runs on the physical architecture to produce f • Rationality ≠ Omniscience, Perfection, Success • Rationality � exploration, learning, autonomy 3 4 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart Bounded Rationality Other AI courses • What if the best strategy given past • CS497: Frontiers of Computer Science percepts cannot be implemented with • CS498: Foundations of Machine Learning today’s computers? • CS498: Image and vision computing • We have seen many theories for • CS785: Intelligent Computer Interaction rational agents but what if those • CS886: Reasoning under Uncertainty theories are intractable? • CS886: Knowledge representation • Bounded rationality: find best • CS886: Theoretical Foundations of clustering implementable strategy given past • CS886: Game-Theoretic Methods in Computer Science percepts 5 6 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart 1
CS497: Frontiers of Computer CS498: Foundations of Machine Science Learning • Instructor: Shai Ben David • Instructor: Shai Ben David • Term: Winter 2007 • Term: Winter 2007 • Objectives: • Objectives: – The purpose of this course is to expose – The course is aimed to familiarize the students undergraduate students to research. It is with the basic theoretical tools and issues meant to be a stimulating course where underlying some of the most useful machine students get a taste of various research learning techniques. The theory of machine topics as well as the opportunity to learning draws from several established explore in more depth a topic of their mathematical areas including statistics, geometry, choice. combinatorics, and computational complexity. 7 8 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart CS785 Intelligent Computer CS498: Image and Computer Vision Interaction • Instructor: Robin Cohen • Instructor: Richard Mann • Term: ??? • Term: ??? • Topics: – multiagent systems, • Topics: TBA – intelligent tutoring systems and knowledge-based systems, – datamining, – user modeling, – natural language generation and dialogue, – plan recognition 9 10 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart CS886: Topics in AI: CS886: Topics in AI: Reasoning under Uncertainty Knowledge Representation • Instructor: Pascal Poupart • Instructor: Chrysanne DiMarco • Term: ??? • Objectives: • Term: ??? – This course will focus on the principles of probabilistic reasoning and sequential decision making for a wide range of • Topics: TBA settings including adaptive and multi-agent systems. The modeling techniques that will be covered are quite versatile and can be used to tackle a wide range of problems in many fields including robotics (e.g., mobile robot navigation, control), computer systems (e.g., autonomic computing, query optimization), human-computer interaction (e.g., spoken dialog systems, user modeling), bioinformatics (e.g., gene sequencing, design of experiments), operations research (e.g., resource allocation, maintenance scheduling, planning), etc. Hence, the course should be of interest to a wide audience beyond artificial intelligence. 11 12 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart 2
CS886: Topics in AI: CS886: Topics in AI: Game-Theoretic Theoretical Foundations of Methods in Computer Science clustering • Instructor: Shai Ben David • Instructor: Kate Larson • Term: Fall 2006 • Term: Fall 2006 • Topics: TBA • Topics: TBA 13 14 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart AI research group My research projects • http://www.cs.uwaterloo.ca/~ppoupart/projects.html • Web: ai.uwaterloo.ca • Professors: • Partially observable Markov decision processes – Shai Ben David (learning theory) • Intelligent assistive technologies – Chrysanne DiMarco (natural language processing) • Spoken dialogue systems – Peter Van Beek (constraint programming) • Bayesian reinforcement learning – Robin Cohen (multi-agent systems, user modeling) • Trust modeling in electronic markets – Pascal Poupart (reasoning under uncertainty, • Ontology learning machine learning, natural language processing) • Preference elicitation – Kate Larson (game theory, mechanism design) – Richard Mann (computational vision) 15 16 CS486/686 Lecture Slides (c) 2006 P. Poupart CS486/686 Lecture Slides (c) 2006 P. Poupart AI jobs • Very few “AI companies” • AI tends to be embedded in many applications • Many companies have AI R&D groups – Intel, Microsoft, IBM, Google, NEC, Yahoo, HP • AI is a growing industry • Has the potential to revolutionize the computer industry! 17 CS486/686 Lecture Slides (c) 2006 P. Poupart 3
Recommend
More recommend
