Automate Hyper-Parameter Optimisation

18 / 45 Training This practice was ranked as advanced. Click to read more.


Intent

Enhance experimentation, performance and fair comparisons between algorithms, by automating hyper-parameter search and model selection.

Motivation

Finding the right hyper-parameters for a model, or choosing between different machine learning models can be a daunting task. Automated methods to perform these activities are now available, with great 'off the shelf' tool support.

Applicability

Automatic hyper-parameter optimisation should be considered in any machine learning application.

Description

The performance of machine learning models depends on the choice of hyper-parameters. Moreover, in many cases one would train different machine learning models (e.g. SVMs or Gradient Boosting), and choose the better performing one.

Instead of manually trying out hyper-parameters or performing manual model selection, one can automate these tasks and gain experimentation speed and performance.

However, it is still recommended that models are peer-reviewed and assessed by team members before deployment to production.

Adoption

Related



18 / 45 Training This practice was ranked as advanced. Click to read more.