Tuesday, May 13, 2014

Personal Notes On Artificial Intelligence [Unofficial]

  • Search
    • Goal State <> Current State
  • Informed Search
    • Choosing next state based on distance from goal state.
      • Greedy Best First Search
        • Distance of next state from goal state
      • A* Search
        • Distance to next state + Distance of next state from goal state
  • Optimization
    • Searching for maximum / minimum.
    • Modifying the current state and evaluating to see if it’s comparatively optimum than the current state.

    • Algorithms

      Hill Climbing

      Simulated Annealing 


    • Local Beam Search
      • Keeping “a number of” “current states” in memory
      • Algorithms

      • Genetic Algorithm
  • Constraint Satisfaction Search
    • Search + Constraint Propagation
  • Planning
    • Search utilizing concepts from logic.
      • logical AND
        • => decomposition of problem
  • Uncertain / Probabilistic Reasoning
  • Probabilistic Reasoning over time
    • Models
      • Hidden Markov Model
      • Dynamic Bayesian Network
  • Simple Decision Problems
    • Decision Theory = Probability Theory + Utility Theory
  • Complex Decision Problems
    • Game Theory
    • Mechanism Design
  • Reinforcement Learning
    • Inductive / Statistical Learning: Learning input/output pair for a particular problem.
    • Knowledge based learning: Adding to what you already know as you go along learning new things.
    • Reinforcement Learning: Learning sequence of behaviors from feedback.
  • Communicating with real world
    • Natural Language Processing
      • Parsing, Semantics, etc.
    • Statistical Language Processing
      • Probabilistic Language Models
        • Counting occurrence of words, N-gram models.
      • Information Retrieval, Information Extraction, Machine Translation.
    • Perception
      • Computer Vision
    • Robotics
      • Sensors
        • Localization
        • Mapping
      • Actuators
        • Degrees of freedom
      • Software Architecture
        • Reactive Architecture
        • Probabilistic Robotics [3]



References

  1. Overview of (Artificially) Intelligent Agents
  2. Artificial Intelligence: A Modern Approach
  3. Probabilistic Robotics (Intelligent Robotics and Autonomous Agents series) by Sebastian Thrun , Wolfram Burgard, Dieter Fox

No comments:

Post a Comment