IAC Winter School 2018: unsupervised machine learning algorithms in Astronomy

The XXX Winter School of Astrophysics on Big Data in Astronomy: http://www.iac.es/winterschool/2018/.

This repository contains the slides from the lectures I gave during the school, and python tutorials. The topics I covered in the school:

1. Introduction to Unsupervised Learning: The main differences between supervised and unsupervised learning algorithms, and the basic anatomy of unsupervised learning algorithms.
2. Clustering Algorithms: K-means, Hierarchical Clustering, and Gaussian mixture models.
3. Decision Trees and Random Forests:: decision trees and their advantages and disadvantages, ensemble methods, random forest, probabilistic random forest, unsupervised random forest for distance assignment.
4. Dimensionality Reduction Algorithms:: principle component analysis (PCA), independent component analysis (ICA), non-negative matrix factorization (NNMF), t-distributed stochastic neighbor embedding (tSNE), uniform manifold approximation and projection (UMAP),auto-encoders, self-organizing maps (SOM), and PINK. Specifically, we discussed methods with which we can interpret the low-dimensional output of such algorithms.
5. Outlier Detection Algorithms: anomaly detection with supervised learning algorithms, isolation forests, and unsupervised random forests for distance estimation.

Related Github repositories:

1. Neural Networks and Deep Learning tutorials by M. Huertas Company: https://github.com/mhuertascompany/IAC_XXX_WINTER/tree/master/morphology
2. Unsupervised Random Forest and distance assignment by D. Baron: https://github.com/dalya/WeirdestGalaxies
3. Probabilistic Random Forest by I. Reis: https://github.com/ireis/PRF

Tutorial Requirements:

1. Anaconda Python 2.7 version: https://www.anaconda.com/download
2. The tutorials are available through Jupyter notebooks. The tutorials use the following packages: numpy, scipy, scikit-learn, and matplotlib.