SCXT 350

Exam #3 Review

Disclaimer: I have attempted to be comprehensive in the following, but important items may have been omitted by mistake. If you see such an omission, please let me know, but you are responsible for all of the lecture material to date.

The third  hour exam will be held on Friday, April 29. It will cover

Be able to
  1. Knowledge Representation: 
    • Frames (as, for example, in CLIPS)
    • Semantic Networks (including ISA networks - brief sketches)
    • Scripts
  2. Expert Systems
    • Architecture of an expert system
    • Simple templates, facts and rules (define and say what they mean)
  3. Connectionism
    • Be able to describe how connectionism differs from symbolic AI as an approach to the problem of building intelligent machines.
    • Be able to construct a simple perceptron, and sketch the corresponding line.  Be able to state the perceptron learning theorem, and to give an example of a problem that can not be solved by a perceptron (and why it can not be solved by a perceptron).
    • Be able to say how a general feed-forward neural network works (Friday's lecture)

There may be more, but I think that's it. If you have any questions, please ask.