25443: Neural Networks
Course Name: Neural Networks
Course Number: 25443
Prerequisite(s): 25411 (Linear Control Systems)
Co-requisite(s): -
Units: 3
Level: Postgraduate
Last Revision: Fall 2012

Description: 
This course will introduce the fundamentals of designing intelligent computer systems. By the end of this course, the students will be able to build agents that efficiently make decisions in fully informed, partially observable and adversarial settings. The techniques learned in this course apply to a wide variety of artificial intelligence problems.
 
Syllabus:
  • Introduction to AI
    • What AI is, history, what AI can do, applications
  • Intelligent agents
    • Agents and environments, concept of rationality, structure of agents
  • Search
    • Searching for solutions, uninformed search methods, breadth-first search, depth-first search, iterative deepening, uniform-cost search
  • Heuristics
    • Informed (heuristic) search methods, greedy search, A* search, heuristic functions.
  • Constraint satisfaction problems
    • Varieties of CSPs, constraint graphs, backtracking search for CSPs, forward checking, constraint propagation, arc-consistency, structure of problems, tree-structured problems, local search methods, simulated annealing
  • Games
    • Deterministic games, minimax search, alpha-beta pruning, examples
  • Utility theory
    • Utility functions, maximum expected utility, expectimax search
  • Markov decision processes
    • Utilities of sequences, MDP search trees, Bellman equations, value iteration, policy iteration
  • Reinforcement learning
    • Passive learning, model-based learning, model-free learning, policy evaluation, active learning, exploration, Q-value iteration, Q-learning
  • Bayes’ nets
    • Probabilistic models, semantics of Bayes’s nets, independence, conditional independence, causal chain, reachability, exact inference, approximate inference, inference by enumeration, variable elimination, prior sampling, rejection sampling, likelihood weighting
  • Probabilistic reasoning over time
    • Inference in temporal models, hidden Markov models, Particle filtering, Kalman filters, dynamic Bayes’ nets, speech and language, HMM for speech recognition
  • Naïve Bayes
    • General naïve Bayes, naïve Bayes training, inference for naïve Bayes, Laplace smoothing, classification, generative vs. discriminative
  • Learning probabilistic models
    • Learning with complete data, learning with hidden variables, EM algorithm
  • Learning from examples
    • Learning decision trees, evaluating and choosing the best hypothesis

References:
  • L. Fausell, Fundamentals of Neural Networks: Architectures, Algorithms, and Applications
  • D. Graupe, Principles of Artificial Neural Networks, 2007
  • S. Haykin, Neural Networks: A Comprehensive Foundation, 2008

 
Last Update: 2024-07-10