Frank C Langbein
Ex Tenebris Scientia
I. Intelligent Systems
In this section we give an overview of different approaches towards artificial intelligence and discuss some attempts to define intelligence. Furthermore, we present a general framework for studying artificial intelligence.
II. Problem Solving by Searching
Relatively simple search algorithms can be utilised to solve many problems for which some assert that they require intelligence for solving. We review standard, uninformed graph search algorithms and discuss in detail informed search algorithms, algorithms for playing two player board games, and constraint satisfaction problem solvers.
III. Knowledge and Reasoning
The search algorithms from II represent the problem they aim to solve in code and are essentially designed to solve a single problem. In this section we will investigate options for representing problems in a general, mathematical language and devise algorithms which can find solutions to problems presented in this language. In other words, this section is all about logic and proofs.
IV. Uncertain Knowledge and Reasoning
Logic is about facts being true or false, nothing in between. But in the real world we are hardly ever 100% certain about a fact. In this section we will introduce methods to handle uncertainty in our knowledge and discuss how to use this for reasoning and making decisions. So we will discuss probability theory and Bayesian theory for representing uncertainty and utility theory for making decisions.
V. Machine Learning
II to III is all about algorithms and operations on representations of the real world to make a decision. These operations are predominantly based on symbol manipulation and the environment is represented with a fairly static model. In this section we consider approaches to learning how the environment works and also study non-symbolic approaches towards creating intelligent behaviour and the way the environment works can be detected; depending on how much time we have left.
Cite as Notes, by Frank C Langbein [24/January/2009, 18:34].
[Contact Details]