Machine Learning and Data Mining

Mathematics for Machine Learning

Machine learning is about designing algorithms that automatically extract valuable information from data.

Introduction

At the core of machine learning are data, models, and learning. At a high level, the machine learning pipeline involves representing data as vectors, choosing an appropriate model, and then learning from the data using numerical optimisation methods. The aim is for the model to generalise well to new/unseen data.

In supervised learning, the aim is to develop a predictive model based on labelled data – data that has both the input and output class (the label). Within supervised learning, there are Regression algorithms that make predictions for numerical values, and Classification algorithms that categorise data into different classes.

In Unsupervised Learning, the aim is to discover an internal representation from unlabelled data (data without an output class given).

Other types of machine learning include semi-supervised, self-supervised, and transfer learning.

Overview

Main Topics

1. Regression
2. Classification
3. Tree Learning
4. Kernel Methods
5. Ensemble Learning
6. Neural Networks
7. Unsupervised Learning
8. Learning Theory

Revision

• Maths
  • Calculus
  • Probability
  • Linear Algebra
• Python
  • NumPy
  • matplotlib