PyML is an interactive object oriented framework for machine learning in Python.
PyML has been tested on Mac OS X and Linux. Some components are in C++ so it`s not automatically portable.
Here are some key features of "PyML":
· Classifiers: Support Vector Machines (SVM), nearest neighbor classifiers, ridge regression
· Multi-class methods (one-against-one and one-against-rest)
· Feature selection (filter methods, RFE, multiplicative update)
· Model selection
· Syntax for combining classifiers
· Classifier testing (cross-validation, error rates, ROC curves, statistical test for comparing classifiers)
What`s New in This Release: [ read full changelog ]
· Added wrapper for liblinear linear SVMs. If you only need a linear
· SVM, these solvers offer a very significant speedup for large
· datasets.
Usage:
· SVM(optimizer = `liblinear`, loss = `l2`) for l2 loss SVM
· or SVM(optimizer = `liblinear`, loss = `l1`) for l1 loss SVM
· Added containers.setData.SetData - a dataset container where each
· example is a set of objects.
· Chris Hiszpanski reported an issue and fix for ROC calculation that
· would fail for a corner case.
· When creating a dataset with numeric labels, the `numericLabels`
· keyword argument was not passed on to the Labels constructor when
· creating such a dataset from arrays/lists
· the "stratifiedCV" method of SVR was being called by model
· selection. That was addressed by defining it to be regular
· cross-validation (stratified CV doesn`t make sense for regression).
· Better solution would be to have a separate base class for regression.
· Can create an empty SparseDataSet or VectorDataSet, and then
· populate it on the fly with features (using its ...