Foundations of Machine Learning: From Data to Prediction

Predictive modeling is rapidly becoming a core professional skill — yet the gap between working with data and building reliable machine learning models remains significant for many practitioners. This course bridges that gap, equipping participants with the foundational concepts and hands-on skills needed to design, implement, and evaluate supervised machine learning models using Python and Scikit-learn. From understanding the learning process to deploying a first end-to-end pipeline, participants gain the rigor and intuition that professional machine learning practice demands.

While Foundations of Machine Learning: From Data to Prediction can be taken as a standalone course, it also serves as the foundation of EPFL’s learning pathway in applied machine learning. It is the first of four Microcredentials leading to the Certificate of Open Studies (COS) in Applied Machine Learning https://www.formation-continue-unil-epfl.ch/formation/applied-machine-learning/, providing a flexible pathway toward an EPFL credential in applied machine learning.

Upon successful completion of this course, participants will be able to:

Classify supervised learning problems as regression or classification tasks
Apply feature engineering, feature selection, and preprocessing techniques to construct suitable feature spaces
Explain the concept of generalisation and the role of the train–test split in modeling
Implement a full data-to-predictions workflow using Scikit-learn and multilinear models
Select appropriate evaluation metrics to measure and interpret model performance
Apply machine learning knowledge to diagnose problems and explore solutions

Nongraded

Professionals already using Python libraries for data analysis and visualization who seek to expand their skills toward machine learning; data analysts and quantitative analysts looking to introduce their first machine learning models into their workflow; and software engineers transitioning into the machine learning and AI space whose long-term aim is to build real-world applications.