Machine Learning🔗

FSD311🔗

Statistical foundations are covered in Fundamentals of Decision Making (FSD310).

This class is divided into three modules:

Schedule
21/09 PM	Introduction to Machine Learning](index.md/#introduction-to-machine-learning)	Supervised and unsupervised learning, data preprocessing and ML workflow
22/09 AM	Geometrical approach	Support Vector Machines, the bias/variance tradeoff and a bit of kernel theory.
29/09 AM	Probabilistic approach, pt 1	Naive Bayes classification and Gaussian Processes
29/09 PM	Probabilistic approach, pt 2	Surrogate Modeling and Bayesian Optimization
30/09 AM	Committee learning, pt 1	Decision Trees and Boosting
06/10 AM	Committee learning, pt 2	Bagging, Random Forests
06/10 PM	Committee learning, pt 3	Anomaly Detection
07/10 AM	Explainability	LIME, SHAP

Schedule
13/10 AM	Artificial Neural Networks	ANNs, Backpropagation, Stochastic Gradient Descent
13/10 PM	Deep Learning	Layers, Architectures, Training
03/11 AM	Deep Learning for Computer Vision, pt 1	Convolutional Neural Networks
03/11 PM	Deep Learning for Computer Vision, pt 2	CNNs and satellite imagery
10/11 AM	Autoencoders and Self-Supervised Learning	autoencoders, SSL
10/11 PM	Dimensionality Reduction	t-SNE, SAEs
17/11 AM	Image generation, pt 1	VAEs and GANs
17/11 PM	Image generation, pt 2	Diffusion Models
24/11 AM	Deep Learning in Practice	training, debugging, and deploying deep models
24/11 PM	RNNs	Recurrent Neural Networks, LSTM, GRU
30/11 PM	Transformers	the Transformer architecture
01/12 AM	LLMs	Large Language Models
01/12 PM	Retrieval-Augmented Generation	RAG

Schedule
11/01 PM	Introduction and MDPs	RL intuitions, Robotics, Markov Decision Processes
18/01 PM	Bellman Equations and Value Functions	Bellman Equations, Characterizing and Evaluating Policies
25/01 PM	Deep Q-Networks	Value Function Approximation, Deep Q-Networks
01/02 PM	Deep Q-Networks (lab)	Hands-on Implementation of DQN
08/02 PM	Actor-Critic methods	Policy Gradients and Actor-Critic Algorithms