COMP551: Applied Machine Learning

We use this repository to share the code for some of the topics that we cover in the McGill COMP551 class. Note that the recent updates in 2025 include the use of LLM tools to generate comments or parts of the code.

Topics Covered

Maximum Likelihood and Bayesian Inference

• Gaussian (Gaussian.ipynb)
• MLE and Bayesian Inference (MLE_BayesianInference.ipynb)

Linear Classification and Regression

• Naive Bayes (NaiveBayes.ipynb)
• Linear Regression (LinearRegression.ipynb)
• Logistic Regression (LogisticRegression.ipynb)
• Perceptron and Support Vector Machines (Perceptron_and_LinearSVM.ipynb)

Training and Evaluation

• Gradient Descent (GradientDescent.ipynb)
• Regularization (Regularization.ipynb)
• Generalization and Model Selection (ModelSelection.ipynb, CurseOfDimensionality.ipynb)

Deep Networks

• Multilayer Perceptron (MLP.ipynb)
• Automatic Differentiation (MLP.ipynb)
• Deep Learning with Image Data (ConvNets.ipynb)
• Deep Learning with Sequential Data (RNN.ipynb)

Nonparametric Methods

• Nearest Neighbours (KNN.ipynb)
• Decision Trees and Random Forests (DecisionTree.ipynb)

Unsupervised Learning

• Clustering (KMeansClustering.ipynb, EMforGaussianMixture.ipynb)
• Dimensionality Reduction (PCA.ipynb)