Overview

This course will introduce students to the field of machine learning by providing a broad overview of all of the various aspects of a machine learning pipeline, as well as the various types and subfields of machine learning models. We will explain various aspects of the data pipeline, such as what to consider during data collection, how to analyze and interpret your datasets, how to create meaningful visualizations of your data and how to clean and prepare your datasets for training machine learning models. These discussions will also provide students with insights regarding how the various aspects of the data pipeline changes for different types of data, such as tabular, image, text and time series data. Students will then learn about the various subfields of machine learning, with a particular focus on the most popular supervised and unsupervised machine learning models, as well as a few deep learning architectures. We will also discuss semi-supervised and reinforcement learning to a lesser extent. Lectures regarding specific models will aim to teach students what the core idea behind the models are, what the main differences between the various models are and what is considered to be their pros and cons. We will not provide detailed mathematical explanations regarding these models, but certain discussions provide some insights into aspects of the underlying mathematics that influence how the models work and what problems they are suitable for.Apart from discussing data pipelines and the various types of machine learning models, this course will also provide students with the necessary information to be able to build their own machine learning solutions for basic to intermediate problems. This includes discussions of the popular machine learning frameworks in Python (Scikit-learn, PyTorch, Tensorflow and Jax), the steps that should be considered when designing a machine learning project, how to train, finetune and evaluate machine learning models in a way that will provide robust performance estimations as well as a few practical examples where machine learning models are applied to some demonstrative datasets.There is considerable overlap between our Introduction to Machine Learning course and this course, but we discuss the various topics in more detail in this course with the aim to enable students to be able to implement their own machine learning solutions by the end of the course.

Skills

Reviews