- Exploratory Data Analysis with Python (missing values, Correlation analysis ...)
- Forecasting with Prophet
Libraries used: numpy, pandas, matplotlib, seaborn, statsmodel, fbprophet (Facebook).
- Project Structure Overview
- Business Context
- Problem Statement
- Desired Outcome
- Version
- Author
- References
├── Time Series Analysis Retail Stores Sales
| ├── data - this folder contains training data.
| │ ├── sales.csv - trimmed data is uploaded due to max limit
| │ └── store.csv
| │
└───── Retail Stores Sales Prediction.ipynb
A large grocery retailer (Shop@ABC) operates multiple stores in the US.
Demand planning for the stores is a crucial element to ensure the stores are not overstocked or under-stocked, thus enabling a competitive edge in the market. However, the business currently believe that they are not able to leverage the power of data driven analytics solution for the same.
The store managers needs to be empowered with data analytics solutions/ tools, which enables them to predict daily sales for up to a month in advance.
Note that the store sales are impacted by many factors, including various types of promotions, holiday seasons, and various store demographics, among others.
They would like to understand various factors that affects the demand of the stores and also develop accurate demand forecasting models for the stores
Using the attached sample data, please come up with your analytics approach:
- Understanding of the different data fields
- Data cleaning/ manipulations and business/ technical logics involved
- Analytical findings and insights (EDA)
- Forecast model development and performance measures
- How would you measure the improvement driven by the solution provided?
- Suggested Next steps
1.0.0
- Rahul Gaikwad - Initial work and development
I welcome your questions. Write to [email protected]