HANDS-ON-LAB

Walmart Sales Prediction Machine Learning Project

Problem Statement

Predict the sales of Walmart stores.

Dataset

Kindly download the data from here.

Tasks

  1. Hypothesis based EDA:

    • Share of sales and profit by product category (top 5)

    • Share of profit by region

    • Share of sales and profit by product name (top 5) 

  1. Plot the trend of sales and profit over time as line chart

  2. Create new features:

    • Create average sales per category and average sales by region features and drop the category and region column

    • One-hot encode all the other categorical variables.

    • Standardize the numerical variables

  1. Build Models and compare the results:

    • Build a Linear Regression and Random Forest on the above data

    • Set the date as index and use only sales column to build ARIMA and SARIMA model

    • Use facebook prophet and build the model

 

Download the dataset and uncover Walmart store sales insights!

 

FAQs

Q1. How can I analyze the sales and profit distribution by product category?

Perform hypothesis-based EDA to identify the top 5 product categories' sales and profit shares.

Q2. How can I determine the profit distribution by region?

Conduct an analysis to uncover the share of profit attributed to different regions.

Q3. Which models can I build for sales prediction?

Use Linear Regression, Random Forest, ARIMA, SARIMA, and Facebook Prophet models to predict Walmart store sales.