Nov. 19, 2001

Noted by Kyuhan Kim

Review / Planning of Lecture

robotics

perception

Artificial Intelligence

Computer Science

Action

Vision

  

Mental Model

Analogical Reasoning

Mental Model

: Generalized representation of line, square, causuality, teleology

(Teleology : intention, purposes , function)

** Model of time

  1. difficulty :

    1. Inference (What is the inference you want to draw? – (because we cannot test it)

      2.  

       

       

       

    2. Not "How do we represent the time?", but what do we use the represenation of time for?

    3. Given Model (Representation of Model of Time), How do we inference from them?

  2. Possible Answer to above questions è Relationship between two events

    1. Instantaneous (Before / After, Temporal Location)

    2. Duration (Start / End, Overlap: temporal overlap)

    3. Periodic (Recurrence)

  3. TimeLines ( one model of time)

    1. Level of Aggregation

    2. Aggregate and Expand (It is similar with Zoom in / Zoom out in robotics)

    3. The problem of this model

      1. Complexity

      2. Slow (Hard to find the right info)

      3. As the times goes on, mapping is very complex

 

 

 

 

 

 

 

 

** Model of Space

  1. Quadtree (one of the models)

    2.  

     

     

     

     

     

     

     

     

     

     

     

     

  2. What is the problem for representing the space ?

  3. What kind of inference can we draw ?

    1. what kind of representaions enable the inference of find, location, inspection, and overlap? How can we represent the inference ?

      • Visual method with multiple resolution

      ..

      ..

      T-rex1

      Next two door, on door poster

      T-rex2

      Next two door,..

      T-rex3

      …

      * Comprehension à several meanings of Understanding in theory.

      1. Robotics / Vision : the Ability to discriminate between two things

      2. Semantic Network : On input à activate other thing

      3. Mental model : Draw the right inference.

      Causuality

      1. What is the causuality ?

        1. A cause B

        2. A is correlated B

        3. A cause B with Probability P

        4. A and B mutually depends on each other

        5. A cause both B and C

        6. A cause B and C occur simultaneously

      2. How could represent the causality?

        1. Causuality is a series of states and transitions.

       

       

       

       

       

      Analogical Reasoning

       

       

       

      ** How can we accomplish with them?

      Water-Flow (Complete)

      Cause

      Greater Flow (beaker, vial, water, pipe)

      Pressure Pressure

      (Beaker) (Vial)

      Liquid (Water) (Supplement)

      FLAT-TOP(Water)

      GREATER

      Diameter Diameter

      Greater (Partial Explanatory)

      Temperature Temperature

      (Coffee) (ice-Coffee)

      ** Relationship Transfer

      : How do we construct the partial explanatory?

      * Difficulty

        1. How do we match them?

        2. How much do we transfer?

       

      Next time to be continued