Page 51 - DCAP310_INTRODUCTION_TO_ARTIFICIAL_INTELLIGENCE_AND_EXPERT_SYSTEMS
P. 51

Unit 3: Representation of Knowledge




               (iii)  The semantic part – that establishes a way of associating real world meanings with  Notes
                    the representations.
               (iv)  The procedural part – that specifies the access procedures that enables ways of
                    creating and modifying representations and answering questions using them, i.e.
                    how we generate and compute things with the representation.

          VII. Knowledge Representation in Natural Language
               Advantages of natural language are:
               (i)  It is extremely expressive – we can express virtually everything in natural language
                    (real world situations, pictures, symbols, ideas, emotions, reasoning).
               (ii)  Most humans use it most of the time as their knowledge representation of choice.
               Disadvantages of natural language are:
               (i)  Both the syntax and semantics are very complex and not fully understood.

               (ii)  There is little uniformity in the structure of sentences.
               (iii)  It is often ambiguous – in fact, it is usually ambiguous.

          3.4.1 Types of Learning

          1.   Perceptual Learning: Ability to learn to recognize stimuli that have been seen before:
                    Primary function is to identify and categorize objects and situations

                    Changes within the sensory systems of the brain
          2.   Stimulus-response Learning: Ability to learn to perform a particular behavior when a
               certain stimulus is present

                    Establishment of connections between sensory systems and motor systems
                    Classical conditioning – association between two stimuli
                         Unconditioned Stimulus (US), Unconditioned Response (UR), Conditioned
                         Stimulus (CS), Conditioned Response (CR)

                         Hebb rule – if a synapse repeatedly becomes active at about the same time that
                         the postsynaptic neuron fires, changes will take place in the structure or
                         chemistry of the synapse that will strengthen it
                         Rabbit experiment – tone paired with puff of air
                    Instrumental conditioning – association between a response and a stimulus; allows
                    an organism to adjust its behavior according to the consequences of that behavior
                         Reinforcement – positive and negative
                         Punishment
          3.   Motor Learning: Establishment of changes within the motor system

          4.   Relational Learning: Involves connections between different areas of the association cortex
          5.   Spatial Learning: Involves learning about the relations among many stimuli
          6.   Episodic Learning: Remembering sequences of events that we witness

          7.   Observational Learning: Learning by watching and imitation other people




                                           LOVELY PROFESSIONAL UNIVERSITY                                   45
   46   47   48   49   50   51   52   53   54   55   56