Under construction. This area is a framework being built: most courses have no teaching material yet, and things may change or break. Private self-study. Not an official programme and not recognised by any ministry or school. No qualification, mark or document is issued, and “years 1 to 3” and “advanced” are names of study paths only.

E14 Stochastic and online optimization

Advanced pathE: Proposed extensionsplannedproposed, not in any programmeabout 160 h (planning estimate)

Planned: syllabus only, no teaching material yet. Nothing on this page can be studied; it describes what the course would cover.

In no programme. This is a proposed extension course; the owner has not approved it into a programme.

Proposed by the assistant to fill a gap in the owner's catalogue. It is in no programme until the owner approves it.

Why this course

The theory behind how large models are actually trained.

Prerequisites

Required: A02 Convex & numerical optimization, G03 Graduate probability
Recommended: none
Can be taken at the same time: none
Everything required, transitively (20 courses): A02, B01, B02, B03, B04, B05, B06, B08, B09, B11, B12, B13, B15, B19, B20, F01, F02, F03, G01, G03
Courses that require this one: none

Concept-level units

Units follow the syllabus order, as a recommended sequence rather than a gate.

Stochastic gradient methods
Stochastic approximation
Variance reduction
Online convex optimization
Regret bounds
Bandits first look

Assessable outcome

Prove a convergence or regret bound and identify the assumptions it needs.

Assessment design

deeper problem sets
reconstruction of an unfamiliar argument
advanced source reading
cumulative examination or substantial project

cumulative; uses spaced retrieval

Resources

Every title here is candidate, unverified unless marked verified. Titles were suggested while planning. Nobody has checked that they exist, what they cost or how they are licensed.

Bubeck · Convex Optimization: Algorithms and Complexity (free)
Hazan · Introduction to Online Convex Optimization (free)
Lattimore and Szepesvári · Bandit Algorithms (free)

Content record

Version

0.0.0

Review

no review of any kind

Provenance

PROPOSED by the assistant on 30 Oct 2026 to fill gaps in the brief's catalogue; not in the owner's brief; needs the owner's approval and a curriculum reviewer

Known issues

Proposed course: not counted by any programme audit until the owner approves adding it to an elective pool.

Type

course

Bridges and laboratories

No bridge or laboratory is attached to this course yet.

You and this course

Reading your plan on this device…