Machine Learning and Data Governance

Organisation over 20 weeks

Four 5-week blocks (4 weeks of classes, 1 week of assessment). The two course units are taken unchanged from the original degree: same content, same assessments.

• Blocks 1 and 2 (weeks 1 to 10): APA 413 — Machine Learning
• Blocks 3 and 4 (weeks 11 to 20): DON 414 — Data Engineering and Governance

APA 413 - Machine Learning

To master the machine learning methods that solve most of the problems encountered in organisations and, above all, to know how to evaluate a model honestly before proposing it.

Learning outcomes

• Design and train supervised models suited to a classification or regression problem.
• Implement a rigorous validation protocol and diagnose overfitting and data leakage.
• Apply unsupervised methods to segment data, reduce dimensionality or detect anomalies.
• Interpret a model's results and discuss their usability with a domain expert.

Content

• Statistical learning framework: risk, bias and variance, capacity, regularisation.
• Supervised models: linear and logistic models, trees, tree ensembles, support vector machines, nearest neighbours.
• Evaluation protocols: data splitting, cross-validation, problem-appropriate metrics, imbalanced classes, probability calibration, data leakage.
• Unsupervised learning: clustering, mixture models, principal component analysis and non-linear dimensionality reduction methods.
• Anomaly and fraud detection; time series: forecasting basics.
• Interpretability: feature importance, partial dependence, local explanations.

Assessment: Continuous assessment (40%): marked practical work. Project (60%): end-to-end solution of a problem with a justified evaluation protocol and an interpretation note for a domain expert. Same project and rubric for both cohorts.

DON 414 - Data Engineering and Governance

To make data a controlled asset: without quality data that are documented and lawfully collected, no model is reliable or deployable.

Learning outcomes

• Analyse the available data in relation to a learning objective and diagnose their quality.
• Design a pipeline for collecting, cleaning, annotating and preparing data.
• Apply loi n°2008-12 du 25 janvier 2008 (Personal Data Protection Act) and copyright law to building a training dataset.
• Develop a data governance policy: roles, documentation, access rights, security, location.

Content

• Data sources: internal data, open data, sensors, field-collected data, synthetic data; selection bias.
• Preparation: cleaning, missing values, integration, feature engineering; relational databases and suitable storage formats.
• Annotation: protocols, guidelines, inter-annotator agreement, participatory annotation, annotators' working conditions.
• Dataset documentation: datasheets, provenance, licence, usage limitations.
• Personal data protection: principles of loi n°2008-12 du 25 janvier 2008, role of the CDP, Acte additionnel CEDEAO A/SA.1/01/10 (ECOWAS Supplementary Act), Malabo Convention; anonymisation and pseudonymisation.
• Copyright and training data: issues raised by loi n°2008-09 du 25 janvier 2008 sur le droit d'auteur et les droits voisins (Copyright and Related Rights Act), to be studied on the text itself.
• Governance and sovereignty: roles, life cycle, security of sensitive data, hosting location, reversibility.

Assessment: Continuous assessment (40%): practical work on data preparation and annotation. Project (60%): datasheet for a dataset and governance policy, with an analysis of compliance with loi n°2008-12. Same arrangements for both cohorts.

Career opportunities

Data specialist, data preparation and governance officer, machine learning model designer.

At your own pace: the self-paced pathway

In a class, this certificate takes 20 weeks, on the cohort's dates. On the self-paced pathway, you start whenever you like, online, and progress at your own pace. The price is the same.

Admission

Bachelor's degree in science, computer science or statistics, or equivalent experience assessed on application; practical programming experience and basic statistics. Admission based on application; detailed requirements set by the school.

The credits you earn are recognised without further assessment if you later join the programme they come from, Professional Master's degree in Artificial Intelligence, within a period set by the certificate regulations.