Open source Machine Learning library written in Java
-
Updated
Nov 15, 2022 - Java
Open source Machine Learning library written in Java
Real-time UPI fraud detection in pure Java: point-in-time feature store with payment-network features, GBDT + exact TreeSHAP, online learning from delayed fraud reports, drift detection, champion/challenger retraining, mule-ring detection, REST API and live dashboard. JDK only, written from scratch.
"Foot on Track" is an application designed to monitor running activities in real-time, authenticate users securely, and provide insightful performance analytics using machine learning.
Dependency-free Java 21 companion to the jeffrey.blog article 'Machine Learning for Backend Engineers': from handwritten rules to a generated decision tree to gradient boosting, behind one Clean/Hexagonal port.
An example of how to provide a CD pipeline enabled h2o Machine Learning project Edit
Restaurant revenue forecasting with Apache Spark, Spark MLlib, MongoDB, feature engineering, and Gradient-Boosted Trees.
To associate your repository with the gradient-boosting topic, visit your repo's landing page and select "manage topics."