🧠 Monthly Sales Forecast with XGBoost

This repository contains an analysis of monthly sales data using machine learning and big data processing tools.

🔍 Project Overview

• Forecast future sales using XGBoost regression
• Visualize trends with 3-month moving averages
• Use PySpark for scalable data preprocessing and transformation

📊 Techniques Used

• 3-Month Moving Average: To visualize trends and reduce noise
• XGBoost Regressor: To forecast the next 3 months of sales
• PySpark: For efficient big data preprocessing and table joins
• Matplotlib: For clear, annotated visualization

📁 Files

• forecast.py — Full script for preprocessing, modeling, and plotting
• forecast_xgb_plot.png — Output chart with actual, smoothed, and predicted values
• monthly_sales.csv — Input dataset
• Sales Forecast - Summary.pdf — Single-page summary of the project

🛠 Requirements

• Python 3.8+
• Libraries: xgboost, pandas, matplotlib, scikit-learn, pyspark

Install them using:

pip install -r requirements.txt

👤 Author

Mehmet Ozturk
📅 April 2025