Python Code for: A White-Box Approach to Predicting Petrol Prices in Zambia Using Machine Learning and Explainable Artificial Intelligence

This repository contains the complete Python source code developed for the study entitled "A White-Box Approach to Predicting Petrol Prices in Zambia: Evidence from Machine Learning and Explainable Artificial Intelligence." The code implements the complete analytical workflow presented in the manuscript, including: Data preprocessing and cleaning Feature engineering Temporal train-validation-test data partitioning Development and evaluation of Ridge Regression, Random Forest, Support Vector Regression (SVR), and Long Short-Term Memory (LSTM) models Model performance evaluation using MAE, RMSE, R², and the Diebold–Mariano (DM) test Explainable Artificial Intelligence (XAI) techniques including Ridge coefficient analysis, SHAP (SHapley Additive Explanations), Partial Dependence Plots (PDPs), and temporal SHAP diagnostics Exploratory Data Analysis (EDA) Residual diagnostics Interactive Streamlit decision-support dashboard for petrol price prediction and model interpretation The implementation accompanies the processed research dataset archived separately on Zenodo and is provided to promote transparency, reproducibility, and future research. The software is intended for academic research, educational purposes, and methodological replication.