Linear Regression
The original line of best fit
Machine Learning algorithms can be difficult to understand.
Implementing the algorithms yourself from scratch is the best way of getting to grips with them.
Each of our modules is designed to make it easier for you to implement the algorithms yourself. We provide an appropriate dataset, help you verify that you've successfully implemented the algorithm, and offer hints and tips for the trickiest parts.
Select one of the modules below to get started, or click here to learn more!
Logistic Regression
Interpretable binary classification
Linear Discriminant Analysis
A generative model for classification
K-Means
Clustering 101
Principal Components Analysis
How to make Big Data, Medium Data
Decision Trees
Recursive Partitioning of the feature space
Bagging
Bootstrapped Aggregation - For when one model just doesn't cut it
Random Forests
Interpretable, robust to overfitting, great name - what more could you want?!
Neural Networks
The basis for deep learning
K-Nearest Neighbours
An intuitive non-parametric approach
Kernel Regression
Locally-weighted, non-parametric regression
Bayesian Linear Regression
A natural approach for uncertainty quantification