Sales Forecasting and Predictive Analysis

This project focuses on predicting sales using machine learning models and visualizing the results with Power BI. The goal is to create accurate sales forecasts based on historical advertising spend data.

Project Overview

• Objective: To predict future sales based on ad spend using machine learning models.
• Dataset: Contains advertising spend data for TV, radio, and newspaper, and total sales.
• Model: Time-series forecasting and regression models were used to predict future sales.
• Outcome: The model helps forecast sales based on historical trends, with interactive Power BI dashboards for insights.

Key Files

• sale_forcasting_Raw_data.csv: The raw dataset containing advertising spend and sales data.
• script.ipynb: Jupyter notebook containing the code for data preprocessing, feature engineering, and model training.
• dashboard.pbix: Power BI dashboard showcasing the sales forecasting insights.
• sale_forcasting_power_bi.pdf: PDF export of the Power BI dashboard visualizations.

Results

• Accuracy: The model was able to predict sales with a high level of accuracy.

• Key Insights:

  • TV ad spend has the highest correlation with sales.
  • The model forecasts a steady increase in sales with increased advertising.

• Visualizations: The Power BI dashboard visualizes the relationship between ad spend and sales, and provides future sales predictions.

Instructions

1. Open the script.ipynb file in Jupyter Notebook to explore the data preprocessing and model training steps.
2. The dashboard.pbix can be opened in Power BI Desktop to interact with the sales forecasting dashboard.