In this project, I developed a web-based dashboard using Python Dash to predict the Loss on Ignition (LOI) of a nickel laterite sample, utilizing the concentration of another element as the predictor variable. To achieve this, I first built a robust regression model, experimenting with both linear and polynomial regression techniques, using the powerful scikit-learn library. After fine-tuning the model and ensuring its accuracy, I exported the trained model as a .joblib file, which allowed for efficient storage and integration.

Subsequently, I incorporated the model into a Dash framework, creating an intuitive and user-friendly front-end interface. This dashboard enables users to interact with the model, inputting relevant data to receive LOI predictions in real-time, effectively providing a seamless user experience for accessing complex predictions in an easily interpretable format.