Clustering
An in depth look at clustering, in particular K-mean clustering. Presented by Professor Simon Parsons.
Clustering is an unsupervised machine learning technique that divides unlabelled data into distinct clusters occupying similar regions of features space.
This video looks at K-means clustering in depth, and introduces the concepts of Gaussian mixture models and expectation maximisation.
Don’t worry about understanding all the equations! They are presented for those particularly interested in the mathematical background. We hope you get the core idea behind the approaches though.
Some terms used are:
- Gaussian model: the concept behind a bell-shaped curve, described by a mean (the centre of the peak) and standard deviation (the spread)
- univariate: a type of data which consists of observations on only a single characteristic or feature.
