Machine Learning and Applications

Overview

Advanced introductory course in machine learning covering both causal and non-causal approaches as well as the foundations of deep learning. The course aims to apply machine learning algorithms to different types of data, understand model training and validation issues, and explore modern data-driven methods.

Technologies used

Key concepts covered

• Theoretical and practical framework for machine learning
• History of probability
• Introduction to causal reasoning
• Causal discovery
• Use of causal models on data
• Non-causal machine learning
• Causal machine learning
• Model training and validation
• Data-driven methods
• K-nearest neighbors (KNN)
• Linear regression
• Generalized linear regression
• Support Vector Machines (SVM)
• Gaussian mixtures
• Gaussian processes
• Kernel methods
• Applying models to different types of data
• Introduction to deep learning algorithms
• Final project applied to real-world data

Coursework and evaluated components

• Lab 1 on a causality-oriented problem
• Selection and description of a dataset
• Presentation of a causal method
• Midterm exam
• Final project with presentation