Skip to content

Using Python, Javascript, Postgres and HTML built a retail dashboard containing interactive visualization and forecast generated by Facebook Prophet and other machine learning models. Grocery recommendation system uses k-means clustering and the surprise algorithm

Notifications You must be signed in to change notification settings

beckyydo/retail-machine-learning

Repository files navigation

Retail Machine Learning

Team Members

  • Rebecca Pham
  • May Lacdao
  • Elizabeth Salas Martinez
  • Hanieh Babaee
  • Ronald Clarke

Deploy: https://retail-ml.herokuapp.com/

Objective

For this project, we decided to build on our two previous projects (ETL Project and Project 2). We will be incorporating machine learning algorithms to:

  1. Predict Walmart's stock price
  2. Create a grocery recommendation system
  3. Forecast sales

Dashboard

Sections

Stock Price Prediction

Data: https://www.kaggle.com/aayushkandpal/walmart-inc-stock-data-19722020-latest

Facebook Prophet

  • run stock price vs date
  • run stock volume vs date
  • run (stock price vs volume) vs date
  • graph stock data, prediction (line, area graph)

Stock FB Prophet

Grocery Basket Customer Recommendations

Data: https://www.kaggle.com/psparks/instacart-market-basket-analysis

A. Grocery List Recommendation:

  • Load csv file, build a clustering/classification model
  • Webscrape/API google images, use chrome driver searching the word/food item and just taking the copy link address of the first search results, adding the csv file
  • HTML, search a member number and the product suggestion will load

B. Other Product Recommendations:

  • Additional Feature: “You May Also Like..."
  • Using Surprise Algorithm provide 3 products a user may also like to purhcase base on their order history

Grocery Login Grocery Landing Grocery Cart

Sales Forecasting

Data: https://www.kaggle.com/naresh31/walmart-recruiting-store-sales-forecasting

A. Sales vs. SocioEconomic Target

  • Regression on Sales vs SocioEconomic Variables (No Date)
  • Regression to see if there’s relationship between sales and socioeconomic variables
  • Graph against the variables to see if there’s any relationship

B. Facebook Prophet Sales vs Date

  • If there’s no relationship above, you can use facebook prophet
  • Graph a line/area graph with the sales and predictions

Sales Forecast

About

Using Python, Javascript, Postgres and HTML built a retail dashboard containing interactive visualization and forecast generated by Facebook Prophet and other machine learning models. Grocery recommendation system uses k-means clustering and the surprise algorithm

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published