KBAI 笔记 02 Intro to CS6737

Preview

Course Preview

Class Goals

Learning Goals

Class Outcomes

Outcomes

Class Strategies

Strategies

Introduction to Computational Psychometrics

Raven's Progressive Matrices

In this course, students will design AI agents to solve Raven's Progressive Matrices.

Examples are below:

2x1 Raven’s Progressive Matrix I 2x1 Raven’s Progressive Matrix II 2x1 Raven’s Progressive Matrix III 2x1 Raven’s Progressive Matrix IV

• 2 possible answers for this one.

2x2 Raven’s Progressive Matrix I 2x2 Raven’s Progressive Matrix II 2x2 Raven’s Progressive Matrix III

• 2 possible answers

3x3 Raven’s Progressive Matrix I

3x3 Raven’s Progressive Matrix I 3x3 Raven’s Progressive Matrix II

• answer is 5.

Exercise: What is intelligence?

If AI passed intelligence tests, does that mean AI is intelligent? Not for sure. it just means AI can process the tests and give correct answers. That's it.

Principles of CS7637

7 principles

1. KBAI agents represent and organize knowledge into knowledge structures to guide and support reasoning.
2. Learning in KBAI is often incremental
3. Reasoning in KBAI agents are top-down as well as bottom-up
4. KBAI Agents match methods to tasks
5. KBAI agents use heuristics to find solutions that are good enough, but not necessarily optimal
6. KBAI agents make use of the recurring patterns in the problem they solve
7. the structure of KBAI agents enables learning, reasoning, and memory to support and constrain each other.

Readings

Readings

Understanding the Natural and Artificial Worlds http://courses.washington.edu/thesisd/documents/Kun_Herbert%20Simon_Sciences_of_the_Artificial.pdf