Skip to content

ML-Analytics library

A centralized location for all machine-learning analytics produced by KX

This enlarges the KX ML-Toolkit and provides a package which can be integrated into KX Insights Platform and Microservices.

They include

  • data preprocessing
  • machine-learning models
  • model/function optimization
  • model evaluation

Most of the functionality in the ML-Toolkit is available, including

  • preprocessing functions
  • clustering models
  • timeseries models

The APIs for these have not changed.

Toolkit documentation

The ML-Analytics library adds to the Toolkit

  • implementations of online learning/out-of-core algorithms
  • updated experimental function signatures for various models to improve ease of use of complex algorithms

ML Graph and Pipeline

This function of the ML-Toolkit is omitted to avoid potential conflation between the concept of a pipeline as it pertains to the Stream-Processor and its use within the ML-Toolkit.