ML Web App

# Description

Full-stack machine learning web application integrating a trained model with an interactive user interface. Allows users to upload images, input data or paste text for real-time predictions. Built with Streamlit, includes: input form with validation, result visualization with charts (matplotlib/plotly), prediction history, explainability with SHAP or LIME, and batch mode for processing multiple inputs. The backend loads the exported model (SavedModel/HDF5), runs preprocessing and returns predictions with confidence levels.

$ ./deploy.sh $ git clone

# Key features

$ Interactive web interface for real-time predictions

$ Exported model loading (SavedModel, HDF5)

$ Result visualization with charts and confidence levels

$ Prediction history with local storage

$ SHAP/LIME explainability for predictions

$ Single and batch mode for multiple inputs

# Gallery

Desktop view

Mobile view

# Technologies used

Python TensorFlow / Keras