Hyper parameter Tuning

Hardik Chheda Updated by Hardik Chheda

Hyper-parameter tuning allows users to tweak their algorithms and alter them to receive the best results for the predictive insights based on the models. 

Below are two ways of optimizing parameters and updating models. 

  • Grid Search, is applied across machine learning to calculate the best parameters to use for any given model. It will automatically scan through data and provide the best optimized model with the most accurate results. It is guaranteed to find the optimal combination of parameters supplied.
  • Random Search is an alternative method of tuning parameters in models, based on random selection of data, subset. This will not completely assure that the result return is the best and most optimized model, since it did not go through all the combination, rather just a random selection. 

In Tellius, we can perform Grid Search and tune the parameters both in AutoML and Point-N-Click.

Note: Tellius does not support Random Search as it does not provide the best results.


When users use AutoML to create predictive insights, Tellius automatically selects a range of parameters that are suitable for algorithms. These parameters make sure that user gets best possible models on their data.

Every algorithm in Machine Language comes with wide variety of parameters to tune. Tuning them manually is highly time consuming. AutoML tries to automate most of this work automated.


In point n click mode, users can select the and set parameters for the each of the models they want to run. They can manually enter values of their choice to optimize the performance of the model and receive more accurate results.


Only Admin Users have the ability to change any preferences for hyper-parameter optimization in the settings.

Once the Admin enables the hyper-parameter optimization, other users will have the best optimized result when using AutoML to perform predictive analysis.

Did we help you?

Model Use