16.412J/6.834J Intelligent Embedded SystemsDescription:Algorithms and paradigms for developing embedded systems that are able to operate autonomously for years at a time within harsh and uncertain environments. Focus on systems thatdemonstrate high levels of deduction and adaptation. Draws upon a diverse set of computational methods from artificial intelligence, operations research, software engineering and control. Topics include: automated mission planning and scheduling, dynamic execution and reactive planning, model-based diagnosis and failure recovery, reinforcement learning, decision theoretic planning, real-time propositional inference, Bayesian inference, and state estimation based on hidden Markov models.Prerequisites: 6.034 or 16.410, 6.041 or permission of instructor. Programming proficiency assumed.3-0-9 H-LEVEL Grad Credit.Lecture: MW 11:00-12:30Location: Rm 33-418Instructor:Brian WilliamsOffice: Space Systems Lab Artificial Intelligence Lab33-330 NE43-838Phone: (617) 253-1678 (617) 253-2739Email: [email protected] hrs by appointmentMailing list: [email protected] [email protected] page: www.ai.mit.edu/courses/16.412J/ (to be posted)www.ai.mit.edu/courses/6.834J/ (to be posted)Readings: - ~ 2 papers from the literature per lecture, or equivalent from "AI a Modern Approach" by Russell and Norvig.Assignments: - ~ 5 Problem Sets- Weekly thought questions- Group lecture on advanced topic (45 minute Presentation, Overview article, Demo).- Final Project: (Proposal, Presentation, Final
View Full Document