Cluster Analysis

New in v2.0: Predictions and find similarities are available

Cluster analysis is a task of grouping similar objects into a number of groups called clusters. It is widely used in market research, medicine, biology, finance, computer science, and more.

Cluster Analysis app will help you to classify your data from any standard or custom object, analyze groups and similarities, and recognize patterns. It uses unsupervised machine learning algorithms, such as K-Means and K-Medoids, to group the records into clusters. These algorithms don't require any training datasets.

With the Cluster Analysis app you can find similar records and predict field values based on configured cluster models. Predictions and similar records and calculated using K-Nearest Neighbors ML algorithm and clustering results.

Cluster Analysis app will also build an interactive visual map of your data using t-SNE dimensionality reduction technique.

The results are stored in a managed custom object so you can easily build a report or export them to a csv file. Optionally cluster numbers can be populated into a source record's custom field.

GitHub: https://github.com/SalesforceLabs/ClusterAnalysis