ELKI supports many clustering algorithms. A few are listed as follows:
- Affinity propagation clustering algorithm: This is a cluster analysis that uses affinity propagation.
- DBSCAN: This is density based clustering especially for the applications with noise; it finds the sets in the database on the basis of density.
- EM: This algorithm creates clusters based on the expectation maximization algorithm.
- AGNES: Hierarchical agglomerative clustering (HAC), or agglomerative nesting (AGNES), is a classic hierarchical clustering algorithm.
- SLINK: This is the single link algorithm.
- CLINK: This is used for complete linkage.
- HDBSCAN: This is an extracting cluster hierarchy.
Also, KMeansSort, KMeansCompare, KMedianLloyd, KMediodsEM, KMeansBisecting, and so on, are some examples from the family of KMean.
A detailed list of clustering algorithms, with all of the algorithms supported by ELKI, can be found at https://elki-project.github.io/algorithms/.
We need to get the required .jar file from https://elki-project.github.io/releases/. Download the executable archive, and download the mouse dataset from https://elki-project.github.io/datasets/.
From the Terminal or Command Prompt, run the following command:
$ java -jar elki-bundle-0.7.1.jar
The preceding command generates the following output:
We can see two options, in an orange color: dbc.in and algorithm. We need to specify the value. In dbc.in, click on the dots (...) to select the mouse.csv file that we downloaded. In algorithm, select k-Mean Clustering algorithm by clicking on the plus sign (+), find kmean.k, and fill it with the value 3. Click on the Run Task button, which is now enabled. It will generate the following output: