Welcome to PyPR’s documentation!

PyPR is an abbreviation for Python Pattern Recognition. The implementation of PyPR was done as part of an Ph. D. project, “Mining of Ship Operation Data for Energy Conservation”, which purpose was to develop statistical models for analyzing the dynamics of ocean-going vessels, especially with regard to modelling fuel effciency. However many of the methods used can be used in a wide range of application, and have therefore been collected in this freely available library (GNU GPL licence). The Ph. D. project was funded by “Føroyagrunnurin frá 1971”, which is administered by Faroese Research Counsil. I am enrolled at DTU Informatics, so the code has been heavily inspired by them, and by the company that I work at, Decision3.

The code is still under constant development, and it might change considerably over time. If you use parts of this code in your project, you might just want to copy the code or files. Please add a reference to this project :)

Contents:

• Getting PyPR
• Preprocessing
  • Normalization
  • PCA
  • Lag Space Matrix
• Clustering
  • K-Means
  • Gaussian Mixture Models
• Artificial Neural Network
  • Network Structure and Evaluation
  • Training
  • Activation and Error Functions
  • Classification
  • Implementation overview
• Gaussian Process
  • Regression
  • Finding the hyper-parameters
  • Application Programming Interface
• Optimization
  • Minimize
• Helpers
  • ANN wrapping
  • Application Programming Interface
• Statistic and Statistical Tests
  • Bayesian Information Criterion
  • Akaike Information Criterion
  • Sample Autocorrelation
  • Ljung-Box Test
  • Box-Pierce Test
  • Application Programming Interface
• More Examples
  • Tapped Delay Neural Network - Sunspot example
  • Gaussian Mixture Model Regression (GMR) - Sunspot example
  • Finding the Number of Clusters to use in a Gaussian Mixture Model
  • Simple GP demo
  • Multiclass classification using ANN
• Project Modules
  • pypr Package

Indices and tables

• Index
• Module Index
• Search Page