The GÉANT project’s Network Development team is pleased to introduce a new learning unit in the AI section of GÉANT’s Network Automation eAcademy: “Unsupervised Learning.”
Part of the “Intelligence Management” track, this course provides a structured introduction to Unsupervised Learning, one of the core paradigms of Machine Learning (ML). Unlike supervised methods, Unsupervised Learning works without labelled targets, allowing algorithms to uncover hidden patterns, groupings, and structures in large datasets. This makes it particularly relevant for network analytics, where labelled data is often scarce or unavailable.
The course is divided into three parts. The first part introduces the principles of Unsupervised Learning, explaining how it differs from Supervised Learning and why it is critical for tasks such as anomaly detection, traffic prediction, and fault diagnosis in computer networks. The second part explores key algorithms, including clustering techniques (K-Means, DBSCAN, Hierarchical Clustering, Gaussian Mixture Models) and dimensionality reduction methods (PCA, t-SNE, and Autoencoders). Learners gain an understanding of how these algorithms function and the types of problems they solve.
The final part focuses on practical application. Through a hands-on Python example with synthetic network traffic data, learners will apply K-Means clustering, use the Elbow Method to choose the number of clusters, visualise results with PCA, and evaluate outcomes using the Silhouette Score. This practical approach enables learners to connect theory with real-world network traffic analysis.
Access the course here: (login required via eduGain or social networks).
Visit our training portal for more information and to access all Network Automation eAcademy units and the interactive training map: .
And don’t miss the chance to join our monthly open meetings. On the first Tuesday of every month, trainers and experts in the Network eAcademy are available during one-hour calls to answer all your questions and collect your feedback and training requests. Simply contact the Network Automation team at network-eacademy@lists.geant.org to receive the link to join the session.
