Artificial Intelligence a modern aproach

Artificial Intelligence a modern aproach

1. What is AI?

   • simulation of human intelligence in machines
   • The Turing Test
     • Proposed by Alan Turing TURING TEST (1950)
     • designed to provide a satisfactory operational definition of intelligence.
   • The computer would need to possess the following capabilities
     • natural language processing
     • knowledge representation
     • automated reasoning
     • machine learning
   • Thinking humanly
     • we must have some way of determining how humans think.
     • cognitive science
       • brings together computer models from AI and experimental techniques from psychology
     • The Greek philosopher Aristotle
       • was one of the first to attempt to codify “right thinking,”
       • provided patterns for argument structures
       • laws of thought were supposed to govern the operation of the mind
       • their study initiated the field called logic
     • In the “laws of thought” approach to AI, the emphasis was on correct inferences.
   • Acting rationally
     • An agent
       • something that acts (agent comes from the Latin agere, to do).
       • A rational agent is one that acts so as to achieve the best outcome
       • The rational-agent approach
         • it is more general than the “laws of thought”
           • the correct inference is just one of several possible mechanisms for achieving rationality.
         • it is more amenable to scientific development than are approaches based on human behavior or human thought

2. The Foundations

   • Philosophy
     • rationalism,
     • dualism.
     • materialism,
     • empiricism
   • Mathematics
     • The first nontrivial algorithm
     • Euclid’s algorithm for computing greatest common divisors
     • incompleteness theorem
       • there are true statements that are undecidable in the sense that they have no proof within the theory
       • which functions are computable
         • Turing also showed that there were some functions that no Turing machine can compute
       • recognize an intractable problem
         • NP-completeness
     • probability.
       • predict the future of an unfinished gambling game
   • Economics
     • Decision theory
       • combines probability theory with utility theory
     • GAME THEORY
       • how to make rational decisions when payoffs from actions are not immediate but instead result from several actions taken in sequence
   • Neuroscience
     • study of the nervous system
     • nerve cells, or neurons
       • mapping between areas of the brain and the parts of the body that they control
   • Computer engineering
     • We need intelligence and an artifact

3. History

   • The gestation
     • Warren McCulloch and Walter Pitts
       • proposed a model of artificial neurons
     • Sources
       • knowledge of the basic physiology
       • function of neurons in the brain
       • formal analysis of propositional logic
       • Turing’s theory of computation
     • HEBBIAN LEARNING
       • remains an influential model to this day.
       • rule for modifying the connection strengths between neurons
       • Donald Hebb (1949)
   • The birth
     • Dartmouth College, which was to become the official birthplace
     • was necessary for AI to become a separate field
     • AI is the only one of these fields that is clearly a branch of computer science
   • great expectations
     • The early years of AI were full of successes
     • Newell and Simon (1976)
       • formulate the famous physical symbol system
         • a physical symbol system has the necessary and SYSTEM sufficient means for general intelligent action
     • Lisp
       • 1958. InMIT AI LabMemo No. 1
       • peapole had the tool he needed, but access to scarce and expensive computing resources
     • limited problems that appeared to MICROWORLD require intelligence to solve
   • Reality
     • early programs knew nothing of their subject matter
     • intractability of many of the problems that AI was attempting to solve
     • The fact that a program can find a solution in principle does not mean that the program contains any of the mechanisms needed to find it in practice.
     • human-level HUMAN-LEVEL AI AI
   • AI becomes an industry
     • return of neural networks (1986–present)
       • the back-propagation learning algorithm
       • Parallel Distributed Processing
     • AI adopts the scientific method 1987
       • approaches based on hiddenMarkov models (HMMs) HIDDEN MARKOV have come to dominate the area
       • AI has finally come firmly under the scientific method.
     • DATA MINING mining technology has spawned a vigorous new industry.

4. What can AI do today?

   • Robotic vehicles
   • Speech recognition
   • Autonomous planning and scheduling:
   • Game playing
   • Spam fighting
   • Logistics planning
   • Robotics:
   • Machine Translation