This documentation is for Machine Learner 1.0.0. View documentation for the latest release.
Skip to end of metadata
Go to start of metadata

The following diagram depicts the collaboration of the main components of WSO2 Machine Learner.

WSO2 ML architecture

The WSO2 Machine Learner architecture comprises of the following main components.

ML Data Layer 

Data layer exposes an interface to ML Core in order to communicate with the underlying ML database.

ML Storage Handler 

Storage handler reads/writes data to a user-configured storage such as a file system, WSO2 Data Analytics Server (DAS), HDFS etc. ML Core uses the storage handler to persist, datasets and generated models.

ML Core 

ML Core is the heart of WSO2 Machine Learner. It handles operations related to the main entities of ML, which are datasets, projects, analyses, and models. Also, ML Core is responsible for creating Apache Spark jobs and submitting those jobs to Apache Spark to perform machine learning analyses.


WSO2 Machine Learner exposes all of its operations via a REST API. This REST API currently supports the basic authentication and cookie based authentication, and the root context is /api. There are five main sub APIs, which are configuration API (/api/configs), datasets API (/api/datasets), projects API (/api/projects), analyses API (/api/analyses), and models API (/api/models).

ML User Interface 

WSO2 Machine Learner provides a Web based user interface to help you with the operations exposed by its REST API.

Database design

The Entity Relationship (ER) diagram below depicts the database design on WSO2 Machine Learner.

ERD diagram

  • No labels