Projects that are tagged with outlier detection.

Logo ELKI 0.7.0-20150828

by erich - September 17, 2015, 10:20:30 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 15584 views, 2841 downloads, 4 subscriptions

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 0.6.0:

  • Uncertain data types, and clustering algorithms for uncertain data.

  • Major refactoring of distances - removal of Distance values and removed support for non-double-valued distance functions. While this reduces the generality of ELKI, we could remove about 2.5% of the codebase by not having to have optimized codepaths for double-distance anymore. Generics for distances were present in almost any distance-based algorithm, and we were also happy to reduce the use of generics this way. Support for non-double-valued distances can trivially be added again, e.g. by adding the specialization one level higher: at the query instead of the distance level, for example.

  • In this process, we also removed the Generics from NumberVector. The object-based get was deprecated for a good reason long ago, and e.g. doubleValue are more efficient (even for non-DoubleVectors).

  • Dropped some long-deprecated classes

Clustering algorithms:


  • speedups for some initialization heuristics
  • K-means++ initialization no longer squares distances (again)
  • farthest-point heuristics now uses minimum instead of sum (renamed)
  • additional evaluation criteria
  • Elkan's and Hamerly's faster k-means variants

CLARA clustering


Hierarchical clustering

  • Renamed naive algorithm to AGNES
  • Anderbergs algorithm (faster than AGNES, slower than SLINK)
  • CLINK for complete linkage clustering in O(n²) time, O(n) memory
  • Simple extraction from HDBSCAN
  • "Optimal" extraction from HDBSCAN
  • HDBSCAN, in two variants

LSDBC clustering

EM clustering was refactored and moved into its own package. The new version is much more extensible.

Parallel computation framework, and some parallelized algorithms

  • Parallel k-means
  • Parallel LOF and variants


  • LibSVM format parser


  • kNN classification (with index acceleration)

Evaluation: Internal cluster evaluation:

  • Silhouette index
  • Simplified Silhouette index (faster)
  • Davis-Bouldin index
  • PBM index
  • Variance-Ratio-Criteria
  • Sum of squared errors
  • C-Index
  • Concordant pair indexes (Gamma, Tau)
  • Different noise handling strategies for internal indexes

Statistical dependence measures:

  • Distance correlation dCor.
  • Hoeffings D.
  • Some divergence / mutual information measures.

Distance functions:

  • Big refactoring.
  • Time series distances refactored, allow variable length series now.
  • Hellinger distance and kernel function.


  • Faster MDS implementation using power iterations.

Indexing improvements:

  • Precomputed distance matrix "index".
  • iDistance index (static only).
  • Inverted-list index for sparse data and cosine/arccosine distance.
  • cover tree index (static only).

Frequent Itemset Mining:

  • Improved APRIORI implementation.
  • FP-Growth added.
  • Eclat (basic version only) added.

Uncertain clustering:

  • Discrete and continuous data models
  • FDBSCAN clustering
  • UKMeans clustering
  • CKMeans clustering
  • Representative Uncertain Clustering (Meta-algorithm)
  • Center-of-mass meta Clustering (allows using other clustering algorithms on uncertain objects) (KDD'14)

Outlier detection changes / smaller improvements:

  • KDEOS outlier detection (SDM14)
  • k-means based outlier detection (distance to centroid) and Silhouette coefficient based approach (which does not work too well on the toy data sets - the lowest silhouette are usually where two clusters touch).
  • bug fix in kNN weight, when distances are tied and kNN yields more than k results.
  • kNN and kNN weight outlier have their k parameter changed: old 2NN outlier is now 1NN outlier, as commonly understood in classification literature (1 nearest neighbor ''other than the query object''; whereas in database literature the 1NN is usually the query object itself). You can get the old result back by decreasing k by one easily.
  • LOCI implementation is now only O(n^3 log n) instead of O(n^4).


  • MiniGUI has two "secret" new options: -minigui.last -minigui.autorun to load the last saved configuration and run it, for convenience.

  • Logging API has been extended, to make logging more convenient in a number of places (saving some lines for progress logging and timing).

Logo Hub Miner 1.1

by nenadtomasev - January 22, 2015, 16:33:51 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 2564 views, 529 downloads, 2 subscriptions

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.

Logo GritBot 2.01

by zenog - September 2, 2011, 14:56:26 CET [ Project Homepage BibTeX Download ] 2992 views, 780 downloads, 1 subscription

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


Initial Announcement on