2.3. Using scikit-learn with Astronomical Data

Machine Learning for Astronomy with scikit-learn

scikit-learn is a Python module integrating classic machine learning algorithms in the tightly-knit world of scientific Python packages (numpy, scipy, matplotlib).

It aims to provide simple and efficient solutions to learning problems that are accessible to everybody and reusable in various contexts: machine-learning as a versatile tool for science and engineering.

Note

This document is meant to be used with scikit-learn version 0.11+ (i.e. the current state of the master branch at the time of writing: 2012-02-13).

• 2.3.1. Astronomy Tutorial setup
  • 2.3.1.1. Setup
• 2.3.2. Machine Learning 101: General Concepts
  • 2.3.2.1. Features and feature extraction
  • 2.3.2.2. Supervised Learning: model.fit(X, y)
  • 2.3.2.3. Unsupervised Learning: model.fit(X)
  • 2.3.2.4. Linearly separable data
  • 2.3.2.5. Training set, test set and overfitting
  • 2.3.2.6. Key takeaway points
• 2.3.3. Machine Learning 102: Practical Advice
  • 2.3.3.1. Bias, Variance, Over-fitting, and Under-fitting
  • 2.3.3.2. Cross-Validation and Testing
  • 2.3.3.3. Learning Curves
  • 2.3.3.4. Summary
• 2.3.4. Classification: Learning Labels of Astronomical Sources
  • 2.3.4.1. Motivation: Why is this Important?
  • 2.3.4.2. Star-Quasar Classification: Naive Bayes
• 2.3.5. Regression: Photometric Redshifts of Galaxies
  • 2.3.5.1. Motivation: Dark Energy, Dark Matter, and the Fate of the Universe
  • 2.3.5.2. A Simple Method: Decision Tree Regression
• 2.3.6. Dimensionality Reduction of Astronomical Spectra
  • 2.3.6.1. SDSS Spectral Data
  • 2.3.6.2. Principal Component Analysis
  • 2.3.6.3. References
• 2.3.7. Exercises: Taking it a step further
  • 2.3.7.1. Exercise 1: Photometric Classification with GMM
  • 2.3.7.2. Exercise 2: Photometric redshifts with Decision Trees
  • 2.3.7.3. Exercise 3: Dimensionality Reduction of Spectra