Frequently Asked Questions

The workshop focuses on clustering methods such as k-means and hierarchical clustering, dimensionality reduction techniques including PCA and t-SNE, and anomaly detection applied to financial datasets. Each session ties the algorithm directly to a practical finance scenario - portfolio grouping, regime detection, or outlier identification in transaction data. Theory is covered only to the depth needed to apply the method correctly.

Participants should have basic Python familiarity and some exposure to financial data - reading a balance sheet or working with price series. No prior machine learning background is required. The workshop builds from first principles, so the math behind each method is explained before the code is written. Participants who have used pandas and numpy before will find the technical ramp-up straightforward.

Each module ends with a hands-on assignment using a real or realistic dataset. Participants submit their notebooks and receive written feedback from an instructor within 48 hours. There is no automated grading - every submission is reviewed by a person. Feedback addresses both the technical approach and the interpretation of results, which is often where the most useful learning happens.

Most participants spend between 6 and 9 hours per week - roughly 3 hours on live or recorded sessions and 3 to 6 hours on assignments and reading. The pace is intentionally moderate so that concepts have time to settle before the next topic is introduced. Participants with demanding work schedules have completed the programme, though some found it necessary to compress their study into two longer days rather than spreading it across the week.

Yes. All live sessions are recorded and available in the participant portal within 24 hours. Recordings remain accessible for 12 months after the workshop ends, so there is no pressure to catch every session in real time. The portal also includes annotated notebooks from each session, which many participants find more useful for review than the recordings themselves.

The workshop uses publicly available financial datasets - equity price series from major indices, anonymised credit data, and macroeconomic time series from sources such as FRED and Yahoo Finance. Participants also have the option to bring their own data for the final project module, provided it does not contain personally identifiable information. Working with familiar data tends to produce more engaged final projects.

Participants who complete all assignments and the final project receive a certificate of completion from jovuloe. The certificate specifies the topics covered and the total hours of study. It is not an accredited qualification, but it documents the specific skills practised during the programme. Several past participants have included it in their professional profiles alongside the GitHub repository of their final project.

Instructors check in with participants who miss two or more submissions. There is a structured catch-up path for each module, and participants can request a one-week extension once during the programme without affecting their completion status. The goal is to finish, not to rush. Falling behind on one module does not mean the rest of the programme is inaccessible - the material stays open throughout.