kabir-kakkar / demand-forecasting

This project focuses on making a Demand Forecasting model. Demand Forecasting is the process of estimating future demand based on historical sales data.

Demand Forecasting

This project is a part of my industrial project at Celebal Technologies.

Note:

This project is not my real work but rather a gist of what I worked on during my internship. The real project is made in Databricks, Azure and Power BI. Due to confidentiality reasons, this has been uploaded.

Time Series Analysis and Forecasting with Prophet

1. Explore the data (ECDF, handle missing values etc).
2. Analysis per store type and correlational analysis of stores activity.
3. Perform extensive Time Series Analysis (seasonal decomposition, trends, autocorrelation).
4. Predict next 6 weeks of sales using Prophet (Facebook methodology).

Libraries used: __numpy, pandas, matplotlib, statsmodel, fbprophet (Facebook)

Power BI Screenshots: