Lecture notes (to appear here incrementally)

Notes 0: Introduction to CSC 244/444
Notes 1: The role of logic & knowledge in AI
Lecture 2: Representing knowledge in FOL
Lecture 3: Semantics of FOL
Lecture 4: Validity and Entailment
Lecture 5: Provability, Soundness and Completeness
Extra material for CSC 444: Natural Deduction
Lecture 6: Resolution Principle Theorem Proving
Lecture 7: Resolution Principle, Continued
Lecture 8: Augmenting Resolution
Lecture 9: Resolution Strategies
Lecture 10 (grads -- optional): Semantic Networks
Slides/Lecture: Natural Logic & Episodic Logic
Slides/Lecture: Specialist Inference
SAT-Solving (Malik & Zhang '09)
SAT-Solving: recent survey (Fichte et al. '23)
Productions Systems (Expert Systems)
Frames
Postscript: Inference with Frames
Nonmonotonic Reasoning
Nonmonotonic Inheritance Networks
To be looked at briefly, for essential ideas:
Default Logic
The following notes (slides) on probability and random variables are for review; you should know the probablity axioms, conditional probability, mutually exclusive (disjoint) events, independent events (and independent random variables), conditional independence, the chain rule, marginalization, and Bayes' rule & its use for inference.
Probability: A Quick Review
The following notes (slides) on boolean Bayesian networks are for review; you should know the form of the joint probability distribution of the boolean (Bernoulli) random variables contained in a BN, the conditional independence assumptions "built into" a BN, the particular relevance of BNs to causal relationships, what subnetwork is relevant to computing a desired conditional probability, how we can compute conditional probabilities by marginalizing over joint probabilities, and "Monte Carlo" (Gibbs sampling) methods of computing conditional probabilities.
Bayesian Networks: A Quick Review
Neural Networks
NN-Based Relation-Inference

the main paper the preceding notes are based on, and a related later paper, are the following:
Freebase inference using neural tensor networks (Socher et al. 2013)
(Optional) Freebase inference with probabilistic NNs (Qu and Tang 2019)
N. Kassner et al., "Language models with rationality"
Planning and the Frame Problem
Silver et al., "PDDL planning with Pretrained LLMs" [PDDL is an extension of STRIPS format]
Pallagani et al., "Plansformer: Generating Symbolic Plans using Transformers" [also uses PDDL]
Drori et al., "A neural network solves, explains, and generates universitymath problems by program synthesis and few-shot learning athuman level"