Course Information

A student who successfully completes COMPSCI.367 will:

• be able to develop computer applications that use Artificial Intelligence techniques
• be able to select and scope a problem for a KBS
• be able to elicit knowledge from domain experts.
• be able to create knowledge level models of domain knowledge
• be able to use a declarative programming language
• be able to use a rule-based knowledge representation and apply inferencing techniques
• have a limited understanding of ontological modeling
• be able to use a knowledge engineering methodology
• be able to use machine learning algorithms and evaluate the statistical significance of their results
• have an understanding of neural networks
• have an understanding of genetic algorithms

In addition, after successfully completing COMPSCI.367 students can

• elicit knowledge and represent it in intermediate knowledge representations
• understand data driven and goal driven inference and can program a declarative rule-based system.
• represent, in a declarative way, what it means for something to be a solution to a given problem
• represent knowledge in predicate calculus and prolog formats
• implement the main heuristic-search-based approaches to problem solving and their pro's and con's
• understand machine learning bias and how that allows programs to learn
• differentiate between induction and deduction with reference to the fundamental importance of bias
• critique stochastic and deterministic search techniques and decide when to use which.
• apply basic stochastic search techniques to solve simple practical problems under uncertainty
• formalize simple search and/or inference problems from real world situations under uncertainty
• critique different meta-heuristic optimization techniques and decide when to use which.
• differentiate between and apply simple linear models
• understand and discuss basic prior and hidden graphical models, e.g. Bayesian networks and Markov random fields.