COMP 456 — Artificial Intelligence
Logic, search, and learning — hand-traceable and exam-ready
A full classical-AI course: propositional and predicate logic, unification and resolution, state-space search (BFS/DFS), heuristic and adversarial search (A*, minimax with alpha-beta), production systems and inference, knowledge representation and expert systems, reasoning under uncertainty, and machine learning from perceptrons to backpropagation. Type the Prolog/Lisp idioms, run the algorithms in Python, and learn to hand-trace every one.
Start course → 0 / 69 lessonsWhat AI Is + How to Study This Course
Logic Foundations: Propositional & Predicate Calculus
- Propositional calculus: symbols, connectives, well-formed formulas
- Truth tables, equivalence, and the logical laws
- From propositional to predicate calculus: terms, predicates, quantifiers
- Semantics: interpretation, model, satisfiable, valid
- Inference rules: modus ponens, modus tollens, instantiation; sound & complete
- Unification, substitution sets, and the most general unifier
- Clause form & Skolemization (bridge to resolution)
Programming AI: Prolog (and Lisp idioms)
- Facts, rules, queries, and the closed-world assumption
- Recursion & backtracking: the Prolog execution model
- List processing: [H|T], member, append, and recursion over lists
- Prolog for graph search: paths and cycle checking
- Collecting solutions: findall, setof, and minimum-cost paths
- Lisp idioms: s-expressions, car/cdr/cons, and recursion
State-Space Search: Graphs, FSMs & BFS/DFS
Heuristic & Adversarial Search
Production Systems, Control & Inference
Knowledge Representation & Expert Systems
Reasoning Under Uncertainty
Machine Learning
- What is learning? Induction, generalization, and inductive bias
- Version space & the candidate-elimination algorithm
- Decision trees & ID3 (information gain)
- Explanation-based learning & analogy (breadth)
- The artificial neuron & the perceptron
- Backpropagation in multilayer networks
- Competitive & associative networks (breadth)
- Genetic algorithms
- Probabilistic ML & the credit-assignment theme (breadth)
Automated Reasoning, NLP & AI as Empirical Enquiry
- Resolution refutation: proving theorems by contradiction
- Horn clauses & the Prolog interpreter
- Weak methods: the Logic Theorist, GPS & means-ends analysis
- Natural language understanding: the pipeline & CFG parsing
- Stochastic NLP: POS tagging, n-grams & PCFGs
- AI as empirical enquiry: PSSH, the paradigms & the generalization problem