Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
tree: 3cef4cb784
Fetching contributors…

Cannot retrieve contributors at this time

45 lines (44 sloc) 1.875 kb
1.7
* New neural network: Hopfield nets added
1.6.1
* Hyperpipes bug fixed
* Data set domains builder returns range for numeric attributes
1.6
* New classifier: Hyperpipes
* New clusterer: Hierarchical Weighted Average Linkage
* New clusterer: Hierarchical Centroid Linkage
* New clusterer: Hierarchical Median Linkage
* New clusterer: Hierarchical Ward's method Linkage
* New clusterer: Diana (Divisive Analysis)
* Example code added: Simple website clusterer
* New simple matching distance function added for discrete attribute vectors
1.5
* ClassifierEvaluator class added: Ideal for experimentation with classifiers
* New classifier: Multilayer Perceptron using Backpropagation neural network
* New clusterer: Hierarchical Single Linkage
* New clusterer: Hierarchical Complete Linkage
* New clusterer: Hierarchical Average Linkage
* Simple Statistics module added
* Simple Proximity functions module added
* New parameter for K-Means based clusterers: Custom centroid function
1.4
* Backpropagation neural networks rebuilt from zero: Faster, leaner code, more parameterizable.
* All algorithms include the Parameterizable module.
1.3
* DataSet class added to wrap data used in classifiers
* All classifiers use the new DataSet class
* All classifiers extend Classifier class
* Clustering algorithm implemented: K Means
* Clustering algorithm implemented: Bisecting K Means
* get_rules method returns generated rules, do not use to_s anymore
1.2
* New module organization
* Added PRISM algorithm implementation
* Added OneR algorithm implementation
* Added ZeroR algorithm implementation
* Example list added to the forrest documentation
* Backpropagation networks: Checking output dimensions during training
1.1
* Backpropagation networks: Train method returns net error.
1.0
* Initial release
Jump to Line
Something went wrong with that request. Please try again.