Teaching

COMPSCI
761

Advanced Topics in Artificial Intelligence, University of Auckland, Semester 2 2022-2023.

A longer course archive covering logic, planning, uncertainty, and reinforcement learning, reorganised in the same visual system as the homepage.

Overview

Course Archive

COMPSCI 761 gathers advanced AI teaching material across symbolic reasoning, planning, decision theory, and modern reinforcement learning. The content here stays faithful to the original course archive while adopting the same editorial structure as the homepage.

Most lectures include public video links together with the corresponding slide deck. Later reinforcement learning content is partly slide-only or mixed-format, matching the original material.

Lectures

Videos & Slides

Lecture 04

Agents Driven by Propositional Logic

Lecture 07

First-order Logic Inference: Logic Programming

Lecture 08

Classical Planning I: Task Representation

Lecture 09

Classical Planning II: Planning via Search

Lecture 10

Classical Planning III: Planning via Inference

Lecture 11

Planning with Uncertainty I: Utilities and Decisions

Lecture 12

Planning with Uncertainty II: Decision Networks

Lecture 13

Reinforcement Learning I: Markov Decision Process

Slides onlySlides
Lecture 14

Reinforcement Learning II: Monte-Carlo Methods

Slides onlySlides
Lecture 15

Reinforcement Learning III: Temporal-Difference Learning

Lecture 16

Reinforcement Learning IV: Value Function Approximation

Lecture 17

Reinforcement Learning V: Policy Gradient Methods

Lecture 18

Reinforcement Learning V: Policy Gradient Methods (Con'd)

Assessment

Tutorials, Assignments & Exam

Assignments & Exam

Tutorials

Tutorial 1

Propositional Logic and Prolog for Assignment 2.

Tutorial 2

PDDL programming for Question 1 in Assignment 3.

Tutorial 3

Reinforcement learning programming for Question 2 in Assignment 3 and final exam Q&A.

Notes

Usage & Copyright

Video access

Video lectures are also publicly available at the University of Auckland library, where copyright applies.

Shared folder

All of the above material can be viewed collectively in this folder.

License and attribution

All material is made available under CC-BY-NC 4.0. Some reinforcement learning slides adapt content from David Silver's lectures. Some third-party content may not be included in the license.