Schedule

Changes to the schedule will be announced on Slack. Lecture PDFs and notebooks are maintained on the course-materials site.

Class record

Date	Lecture	Topic and material
Tuesday, 4 August 2026	L1, part I	Where Deep Learning Losses Come From
Friday, 7 August 2026	L1, part II	Where Deep Learning Losses Come From
Tuesday, 11 August 2026	L1, conclusion	Complete the L1 slides, then run the main notebook in Colab and the robust-regression notebook in Colab. Rendered versions: main · robust.

Course sequence

Module	Lectures	Main topics
Foundations	1–4	Likelihood and losses; neural networks; calculus; backpropagation
Optimization and training	5–7	Optimization, trainability, and regularization
Computer vision	8–11	Convolutional networks, transfer learning, detection, and segmentation
Sequences and language	12–15	Language modelling, recurrent networks, convolutional sequence models, and attention
Transformers	16–17	Self-attention and modern Transformer families
Representation learning	18–19	Self-supervision, vision Transformers, and multimodal models
Generative modelling	20–23	VAEs, GANs, and diffusion models
Systems and synthesis	24–25	Efficient inference and course synthesis

The detailed lecture list is available on the lecture-decks page.