Back to activities

Machine learning crash course


Sep 18, 2018   01:00 PM — 05:15 PM

Maxime Gasse Polytechnique Montréal, Canada

This course intends to provide a general understanding of some of the main topics in the machine learning field. The subjects that will be discussed during the course are listed below:

I. Supervised learning (2h)
• Decision theory (risk minimization, Bayes-optimal prediction)
• Statistical learning (capacity, under/overfitting, bias/variance, regularization)
• Some popular models

    o kNN: k-nearest neighbours
    o CART: classification and regression trees
    o RF: random forests
    o SVM: support vector machines
    o ANN: artificial neural networks

II. Unsupervised learning (30min)
• Some popular problems

    o structure learning
    o outlier detection
    o Clustering
    o manifold learning
    o Sampling
• K-means for clustering
• Gaussian mixtures for clustering (and more)

III. Deep learning (1h30)
• Perceptron era: 1958-60s (biological inspiration, McCuloch-Pitts artificial neuron)
• Backpropagation era: 1986-90s (gradient descent, CNNs, RNNs)
• Deep learning era: 2006- (modern activations, resnets, GANs)


1pm - 3pm: Supervised learning
3pm - 3:15pm: coffee break
3:13 - 5:15 pm: Unsupervised learning + Deep learning


Luciano Costa organizer
Matthieu Gruson organizer


Room 4488
André-Aisenstadt Building
Université de Montréal Campus
2920, chemin de la Tour
Montréal QC H3T 1J4