omniaraouf/hyperparameter-tuning

Hyperparameter Tuning web app made with Streamlit

Hyperparameter Tuning

A Streamlit app for hyperparameter tuning of classifier models using GridSearchCV

• https://share.streamlit.io/randell-janus/hyperparameter-tuning/main/app.py
• Deployed via Streamlit Sharing

About

GridSearchCV

• A function that comes from the Scikit-learn's model_selection package.
• It helps to loop through predefined hyperparameters and fit your estimator (model) on your training set.

Dataset Source

The datasets used are the Iris Plants and Wine Recognition from Scikit-learn's toy datasets.

Models Used

• Random Forest Classifier
• Support Vector Machine
• Logistic Regression

Web App Features

• Switch between two datasets
• Configurable Parameters
  • Random Forest - tune the n_estimators and max_depth.
  • SVM - Specify the kernel types to be included in the tuning and tune C parameter.
  • Logistic Regression - Specify the norm used in the penalization and customize the C parameter.

Views

• Main page
• Full user settings view & Parameter outputs