AutoML Hyperparameter Tuner
AutoML Hyperparameter Tuner is a modular and interactive Streamlit application that automates hyperparameter tuning using Optuna. Built with extensibility in mind, it supports multiple ML models like XGBoost, SVM, and Random Forest, with auto-detection for classification or regression tasks. The project showcases my expertise in Python, machine learning pipelines, and UI-driven development, integrating real-time optimization, model export, and collaborative-friendly deployment.
The AutoML Hyperparameter Tuner is a dynamic, end-to-end Streamlit-based machine learning automation tool built in Python. It simplifies model selection, preprocessing, and hyperparameter optimization for tabular datasets. Designed to support both classification and regression tasks, the project enables users—especially those without deep ML experience—to achieve high-performing models with minimal effort through a clean web UI.
Key Features:
Auto Task & Model Type Detection Automatically identifies whether the dataset requires a regression or classification model based on the target column’s properties, and switches between models accordingly.
Multi-Model Support Implements a flexible backend supporting Random Forest, Logistic Regression, SVM, and XGBoost for both regression and classification, with more models easily extendable.
Optuna-Powered Hyperparameter Optimization Integrates Optuna for efficient hyperparameter tuning using advanced search algorithms like TPE, with Streamlit sliders allowing users to control the number of trials.
CSV Upload & Preprocessing Allows users to upload custom datasets, handles missing values, encodes categorical features, scales numerical values, and manages datetime columns gracefully.
Real-Time Metrics & Visualizations Displays performance metrics like accuracy or RMSE and shows Optuna optimization history plots, helping users track tuning progress visually.
Model Saving & Download Automatically saves the best model using joblib and provides a download button for easy export and reuse.
Robust Error Handling Includes input validation, model fit error handling, and dynamic feedback if the dataset structure is invalid or misconfigured.
Project Gallery
Technologies Used:
Python: Primary programming language used to build the AutoML system, enabling data handling, model training, and integration with Streamlit.
Streamlit: Web framework used to develop an intuitive and interactive user interface for uploading data, selecting models, tuning hyperparameters, and visualizing results.
Scikit-learn: Machine learning library used for implementing classic ML algorithms like Random Forest, Logistic Regression, and SVM, as well as preprocessing pipelines.
Optuna: Advanced hyperparameter optimization library utilized for efficient model tuning via trial-based search strategies.
XGBoost: Gradient boosting library integrated for high-performance model training in both regression and classification tasks.
Pandas & NumPy: Libraries used for data preprocessing, manipulation, and numerical computation within the ML pipeline.
joblib: Serialization library used to persist and export the best-performing models for reuse or deployment.
Git & GitHub: Version control system and code hosting platform employed for collaboration, source tracking, and project deployment.
Demo Link
