ELKI (Environment for Developing KDD-Applications Supported by Index-Structures) is a data mining framework designed primarily for research and teaching. It places particular emphasis on clustering, outlier detection and the efficient processing of spatial and multidimensional datasets.
Unlike general-purpose machine learning environments, ELKI focuses on making algorithms easy to compare and extend. Researchers can combine different data representations, distance functions, indexing methods and algorithms to investigate their behaviour under controlled conditions.
A central design principle is the separation of data management from the mining algorithms themselves. This allows researchers to examine how indexing and other implementation choices affect performance without unnecessarily changing the algorithms being studied.
ELKI contains implementations of many published methods and is particularly useful for benchmarking, evaluating new approaches and experimenting with unsupervised data analysis.
This is free and open source software.
Key Features
- Extensive collection of clustering algorithms, including k-means variants, hierarchical clustering and density-based methods.
- Outlier detection algorithms for identifying anomalous observations.
- Distance-based, density-based and statistical approaches to unsupervised learning.
- Spatial indexing structures including R-tree and R*-tree variants.
- Optimised index structures for accelerating similarity and nearest-neighbour searches.
- Configurable distance and similarity functions for different data types.
- Separation of data management, indexing and mining algorithms.
- Support for custom data parsers and input formats.
- Parameterisation facilities for configuring algorithms and experiments.
- Evaluation measures for comparing clustering and outlier detection results.
- Visualisation of clusters, outlier scores and algorithm behaviour.
- OPTICS reachability plots and other specialised data mining visualisations.
- Support for dimensionality reduction and data transformation methods.
- Benchmarking facilities for comparing algorithms under consistent conditions.
- Extensible architecture for implementing new research algorithms.
- Documentation and publication references describing implemented methods.
Website: elki-project.github.io
Support:
Developer: Ludwig Maximilian University of Munich and contributors
License: GNU Affero General Public License v3.0
ELKI is written in Java. Learn Java with our recommended free books and free tutorials.
Related Software
| Data Mining Software | |
|---|---|
| R | Software environment for statistical computing and graphics |
| MOA | Software environment for data stream mining |
| Orange | Component-based framework for machine learning and data mining |
| astroML | Python module for machine learning and data mining |
| ROOT | Aimed at solving the data analysis challenges of high-energy physics |
| ELKI | Data mining software framework developed for use in research and teaching |
| DataMelt | Full-featured data-analysis framework for scientists, engineers and students |
| KNIME | Konstanz Information Miner |
| Weka | Waikato Environment for Knowledge Analysis |
| RapidMiner | Knowledge discovery in databases, machine learning, and data mining |
| Rattle | Gnome cross platform GUI for Data Mining using R |
Read our verdict in the software roundup.
 LinuxLinks helps you discover the best free and open source software for Linux. Explore our extensive collection of recommended applications, including alternatives to popular proprietary software.We also publish software reviews, Linux tutorials, hardware reviews, free programming books and articles covering the wider open source community. Browse our Linux resources to discover more. Know a useful free and open source Linux application we haven't covered? Suggest a program for inclusion. Has your software been featured on LinuxLinks? Display one of our Featured on LinuxLinks badges on your website or GitHub README. |
Please read our Comment Policy before commenting.