MLPYA free and high-performance Python/NumPy based tool for machine learning | |
Download |
MLPY Ranking & Summary
Advertisement
- License:
- GPL
- Price:
- FREE
- Publisher Name:
- MPBA Group
- Publisher web site:
- http://mpba.fbk.eu/en/home
- Operating Systems:
- Mac OS X
- File Size:
- 149 KB
MLPY Tags
MLPY Description
A free and high-performance Python/NumPy based tool for machine learning Machine Learning PY (MLPY) is an open source and high-performance Python/NumPy based solution for machine learning. MLPY is cross-platform and is available for Mac OS X, Linux and Windows. Here are some key features of "MLPY": Classification: · Support Vector Machines (SVMs) (linear, gaussian, polinomial, terminated ramps) · Spectral Regression Discriminant Analysis (SRDA) · Fisher Discriminant Analysis (FDA) · Penalized Discriminant Analysis (PDA) · Nearest Neighbor (NN) Feature Weigthing: · Support Vector Machines (SVMs) · Iterative Relief (I-RELIEF) · Spectral Regression Discriminant Analysis (SRDA) · Fisher Discriminant Analysis (FDA) · Penalized Discriminant Analysis (PDA) · Discrete Wavelet Transform (DWT) Feature Ranking: · Recursive Feature Elimination (rfe, onerfe, erfe, bisrfe, sqrtrfe) · Recursive Forward Selection (rfs) Resampling Methods: · K-fold · Monte Carlo · Leave-one-out · User-defined split schema · Stratification over labels is also available · Metric Functions Feature List Analysis: · Canberra indicator · Borda count · Landscaping Tools Requirements: · Python 2.4 or later What's New in This Release: New features: · Svm optimal offset option added · FSSun for feature weighting/selection added · Dlda: adaptive offset for classification implemented · Srda: memory usage optimization, speeded up · added Tversky kernel for SVM Bug fixes: · fixed gaussian weights for SVM
MLPY Related Software