skll: SKLL 1.1.0

Implemented enhancements

• Base estimators other than default are not supported for AdaBoost(#238)

• Need cv_folds option(#222)

• Remove decision tree and random forest learners from requires_dense set(#207)

• stratification during cross-validation should be optional (#160)

• Bug when checking if hasher_features is a valid option(#234)

• need to improve the way to check the validity of configuration file(#223)

• Stop modifying global numpy random seed(#220)

• Config file paths are relative to current working directory instead of config file location(#213)

• cannot import name BaseLibLinear(#241)1)

• a line causing an error(#236)

• Changes based on big LibLinear refactoring in scikit-learn(#235)

• skll breaks importing missing class BaseLibLinear with scikit-learn 0.16(#233)

• Learner.model_params will return weights with the wrong sign if sklearn is fixed (#111)

• Overhaul configuration file parsing (@desilinguist, #246)

• Several minor bugfixes (@desilinguist, #245)

• Compatibility with scikit-learn v0.16.1 (@desilinguist, #243)

• Expose cv_folds and stratified (@aoifecahill, #240)

• Adding Report tests (@brianray, #237)

Full Changelog