WSO2 Machine Learner

WSO2 Machine Learner takes data one step further, pairing data gathering and analytics with predictive intelligence: this helps you understand not just the present, but to predict scenarios and generate solutions for the future.

It helps developers and data scientists perform explorative data analyses, generate solutions - called ‘models’ in machine learning - and to compare these models with one another. Integration with WSO2 products like the Complex Event Processor helps you build models for all manner of tasks - such as classifying data, predicting outcomes or detecting anomalies.



Most problems can be solved with clever architecture and code. Some, however, can’t. Some problems - like predicting business scenarios, driving a vehicle, or detecting handwriting - need dynamic, adaptable solutions based on past and present data. This is where Machine Learner comes in.

WSO2 Machine Learner extracts features from a dataset, made available from a file system, HDFS or WSO2 DAS. This data is passed on to the Machine Learner Core, which allows you to explore your datasets, pre-process your data and apply various machine learning algorithms to make sense out of it all. Using Apache Spark, it then analyzes and build models with the algorithms you’ve chosen. Deep Learning provides advanced machine learning capability as image classification; Anomaly detection support helps to detect difficult-to-identify instances of fraud, network intrusions and other disruptions; Recommendation Engine capability allows you to provide product recommendations for users.


All of Machine Learner’s functionality is exposed via a REST API, with a web-based interface to make the functionality available to anyone who wants to access the capabilities of Machine Learner in a more human-friendly manner.


Extract, pre-process, and explore your data

  • Extract data from CSV or TSV file formats, HDFS or WSO2 DAS
  • Use multiple visualizations to explore your data - scatter plots, histograms, Trellis charts, parallel sets and cluster diagrams
  • Use feature engineering to pre-process data for better results
  • Easy graphical user interface for human-friendly viewing

Create models, tune algorithms and make predictions

  • Use Machine Learner to make predictions from generated models
  • Exploit Deep Learning capability leveraging H2O’s Stacked Autoencoders Classifier.
  • Select from machine learning algorithms - Decision Trees, Linear Regression, Lasso Regression, SVM, Naive Bayes for supervised learning and K-Mean clustering for unsupervised learning on your data
  • Employ Anomaly Detection using K Means Algorithm to identify fraud, network penetration and other difficult scenarios
  • Use Machine Learner as Recommendations Engine using Collaborative Filtering Algorithm for predicting customer choices
  • Evaluate models with performance measures like ROC Curve, Confusion Matrix, Accuracy and Root Mean Squared Error
  • Generate, compare, download and export models, with PMML support for interoperability
  • Tune models based on hyper parameters of Iterations, learning Rate and SGD data fraction
  • Enjoy faster model generation for large datasets using in-memory, distributed cluster computing courtesy of Apache Spark
  • Explore feature statistics for shorter training times and easier representation

Integrate for better intelligence

  • Captures data from messages communicated via WSO2 ESB, and passes them to a generated model to carry out predictions
  • Ships with REST APIs that let you control all of the Machine Learner’s functionality - read and modify datasets, analyze and model from afar
  • Integrates with WSO2 Data Analytics Server to form the WSO2 Analytics Platform