LightGBM

This plugin provides LightGBM as a custom model within the DSS Visual Machine Learning interface

Plugin information

Version 1.0.0
Author Dataiku (Du PHAN)
Released 2019-12-03
Last updated 2019-12-03
License MIT License
Source code Github
Reporting issues Github

Description

LightGBM is a gradient boosting framework that uses tree based learning algorithms. It is designed to be distributed and efficient with the following advantages:

  • Faster training speed and higher efficiency.
  • Lower memory usage.
  • Better accuracy.
  • Capable of handling large-scale data

Installation

This plugin does not provide a code environment. You need to run your model training within a code environment where the package lightGBM and jinja2 are installed.
If you want to use other algorithms along side with lightGBM in the Visual Machine section, we recommend that you install the Visual Machine Learning set of packages Administration -> Code envs -> Choose your code env -> Packages to install -> Add sets of packages -> Visual Machine Learning.

How to use

  • In your visual ML task, go to Advanced -> Runtime environment -> Choose the appropriate code environment.
  • Select the LightGBM custom algorithm in your visual machine learning tasks.

Please see the lightGBM document for more details about the algorithm parameters.