Skip to main content

Advanced AI

Block 4.2, 2026

Lecture slides for the Advanced AI major at the University of Limerick.

Lectures

00
Overview Delivery shape, assessment, academic integrity, the text
Week 1 · 7 Sep PDF
01
Supervised Learning Linear models, residuals, least squares, the loss surface
Week 1 · 7 Sep PDF
02
Neural Networks ReLU hinges, hidden units, regions, shallow and deep
Week 1 · 11 Sep PDF
03
Loss Functions Maximum likelihood, log-likelihood, negative log-likelihood, inference
Week 2 · 14 Sep PDF
04
Fitting Models Parameter space, gradients, Jacobian and Hessian, SGD to Adam
Week 2 · 18 Sep PDF
05
Gradients and Initialisation Backpropagation, reverse-mode differentiation, Kaiming/He initialisation
Week 3 · 21 Sep PDF
06
Measuring Performance Generalisation, bias–variance decomposition, double descent, validation and hyperparameter search
Week 3 · 25 Sep PDF
07
Regularisation Explicit and implicit regularisation, early stopping, ensembling, dropout, noise, Bayesian inference, transfer learning and augmentation
Week 4 · 28 Sep PDF
08
Convolutional Networks Invariance and equivariance, convolution in 1D and 2D, padding, stride, dilation, channels, receptive fields, downsampling and upsampling
Week 4 · 2 Oct PDF
09
Residual Networks Degradation, shattered gradients, residual connections, variance growth, batch normalisation, and ResNet, DenseNet and U-Net
Week 5 · 5 Oct PDF
10
Transformers Self-attention as content-dependent routing, the transformer layer, encoder and decoder models, what an attention weight is evidence of, and what a language model's fluency shows about meaning and mind
Week 5 · 9 Oct PDF
11
Generative Models and Reinforcement Learning Softmax with temperature, then two surveys: how GANs, normalising flows, variational autoencoders and diffusion models each learn to generate, and what changes when a model learns from the rewards its own actions earn
Week 6 · 12 Oct PDF
12
Exam Preparation This week's deadlines and venues, the final exam's structure, how it was designed and why, sample questions, and a written question with its model answer and marks
Week 7 · 19 Oct PDF

Labs

* marks a graded notebook, one per week. Submit all of them by Wednesday 21 October, in one zip named Labs_Firstname_Lastname_StudentID.zip, and check it first with check_labs.py. Upload it to Brightspace and email a copy to eoin+advanced-ai-labs@eoin.ai with the subject Advanced AI labs: Firstname Lastname, StudentID.

Supervised learning Prince chapters 1 to 2 · notebooks open in Colab
Week 1 · Mon 7 Sep
Neural networks Prince chapters 3 to 4 · notebooks open in Colab
Week 1 · Fri 11 Sep
Loss functions Prince chapter 5 · notebooks open in Colab
Week 2 · Mon 14 Sep
Fitting models Prince chapter 6 · notebooks open in Colab
Week 2 · Fri 18 Sep
Gradients and initialisation Prince chapter 7 · notebooks open in Colab
Week 3 · Mon 21 Sep
Measuring performance Prince chapter 8 · notebooks open in Colab
Week 3 · Fri 25 Sep
Regularisation Prince chapter 9 · notebooks open in Colab
Week 4 · Mon 28 Sep
Convolutional networks Prince chapter 10 · notebooks open in Colab
Week 4 · Fri 2 Oct
Residual networks Prince chapter 11 · notebooks open in Colab
Week 5 · Mon 5 Oct
Transformers Prince chapter 12 · notebooks open in Colab
Week 5 · Fri 9 Oct
Generative models and reinforcement learning Prince chapters 15 to 19 · notebooks open in Colab
Week 6 · Mon 12 Oct

Project

Project specification What to build, what to submit, and how it is assessed

Quizzes

Sample quiz: week 3 Quiz 3's questions with the answers marked
Sample quiz: week 4 Quiz 4's questions with the answers marked

Key dates

Continuous Assessment Quiz 1
Week 2 · Fri 18 Sep · 09:00–09:10
Continuous Assessment Quiz 2
Week 3 · Fri 25 Sep · 09:00–09:10
Continuous Assessment Quiz 3
Week 4 · Fri 2 Oct · 09:00–09:15
Continuous Assessment Quiz 4
Week 5 · Fri 9 Oct · 09:00–09:15
Final Exam
Week 7 · Tue 20 Oct · 09:00–11:00
Graded Lab Submission
Week 7 · Wed 21 Oct · 12:00
Project Submission
Week 7 · Thu 22 Oct · 12:00
Project Interviews
Week 7 · Fri 23 Oct

Resources

Understanding Deep Learning Simon J. D. Prince · Core text
Mathematics for Machine Learning Deisenroth, Faisal & Ong · Supplementary text
Dive into Deep Learning Zhang, Lipton, Li & Smola · Supplementary text
Essence of Linear Algebra 3Blue1Brown · Video series