About: ELKI is a framework for implementing data-mining algorithms with support for index structures, that includes a wide variety of clustering and outlier detection methods.

Additions and improvements from ELKI 0.7.0 to 0.7.1:

Algorithm additions:

• GriDBSCAN: DBSCAN using grid partitioning (Minkowski distances only)

• Compare-Means and Sort-Means k-means variations (much faster than traditional k-means)

• Visualization of dendrograms.

Important bug fixes:

• Classes with no package ("default package") would cause errors.

• The fast power function implementation was sometimes returning incorrect results.

• Random sampling was sometimes not sampling from the full data set.

UI improvements:

• The file input source will now automatically choose the Arff parser for .arff files.

• MiniGUI now allows choosing other applications.

• MiniGUI now displays the command line in a separate field.

• MiniGUI displays an error message, if an incorrect classpath or JAyatana (on Ubuntu) is detected.

• Export to png now works, we added a work-around for an open Batik bug.

Smaller changes:

• Many smaller bug fixes.

• C-Index for cluster evaluation now can process larger data sets.

• OPTICS output of undefined reachability fixed.

• External distance matrixes are easier to use and perform additional checks.

• Precomputed distance matrixes can answer range and kNN queries.

• Voronoi visualization can be switched in the menu now.

• Improved backwards command line compatibility with additional aliases.

• Added generated @since annotations in JavaDoc.

• Many new unit tests, renamed to the Java conventions.

• Low-level reading of service files, to have faster startup.

About: Hubness-aware Machine Learning for High-dimensional Data

• BibTex support for all algorithm implementations, making all of them easy to reference (via algref package).

• Two more hubness-aware approaches (meta-metric-learning and feature construction)

• An implementation of Hit-Miss networks for analysis.

• Several minor bug fixes.

• The following instance selection methods were added: HMScore, Carving, Iterative Case Filtering, ENRBF.

• The following clustering quality indexes were added: Folkes-Mallows, Calinski-Harabasz, PBM, G+, Tau, Point-Biserial, Hubert's statistic, McClain-Rao, C-root-k.

• Some more experimental scripts have been included.

• Extensions in the estimation of hubness risk.

• Alias and weighted reservoir methods for weight-proportional random selection.

About: GritBot is an data cleaning and outlier/anomaly detection program.

Initial Announcement on mloss.org.