Skip to content

jbdelfau/Drosofhype

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

====================== DROSOFHYPE

Directed, Rapid and Optimized Search Of Fitting HYPErparameters.

Features

A genetic algorithm that searches for the optimal hyperparameters of a machine-learning model.

Quickstart

This library requires a machine-learning model and a json specifying the genetic search configuration. 2 examples are provided in the files test_svm.py and test_xgbregressor.py.

Let's start with the parameters of the genetic search:

  • population_size: size of each generation.
  • n_parents: number of best performing individuals kept to generate children.
  • mutation_rate: occurence of children mutations(between 0 and 1).
  • additional_properties: information that must be included in the output table as additional columns.

In addition, the json must contain an additional key - hyperparameters - describing which parameters of the model have to be varied and in what range. See the above files for more examples.

About

Genetic algorithm for hyperparameters search

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages