The second introductory AI course: trees and graphs, working in notebooks, and a first tour of machine learning and deep learning, from the steps of a project to the main families of algorithms.

Type

Course (part 2 of 3)

Duration

60h - 15 blocs/sittings

Materials

6 slide decks (PDF)

Language

French

View the course on GitHub

What the course covers

The first weeks cover data structures used throughout AI: trees (vocabulary, storage as a list, drawing with Graphviz, preorder, inorder and postorder traversal) and graphs (directed, undirected and weighted adjacency matrices, graph algorithms, and Dijkstra’s algorithm explained in theory, then run in Python with SciPy).

Students then learn to work in notebooks with Google Colab: loading large datasets, showing trees and graphs in cells, plotting, and understanding where a notebook and an IDE each fit in a project.

The last part introduces machine learning and deep learning (with a CNN example on MNIST), the steps of an ML project in professional and academic settings, from exploration to model design and report writing, and the main learning types: supervised classification and regression, unsupervised clustering and dimensionality reduction (PCA), semi-supervised learning and reinforcement learning.

Structure

• Sessions 1–2: trees
• Sessions 3–4: graphs and paths
• Sessions 5–6: notebooks and project
• Sessions 7–8: work periods
• Session 9: introduction to deep learning
• Session 10: steps of an ML/DL project
• Sessions 11–13: machine learning algorithms
• Session 14: work period
• Session 15: final exam

Technologies

Python, Graphviz, SciPy, Google Colab and Jupyter notebooks, scikit-learn, TensorFlow and Keras.

View the course on GitHub