• Zhang, Yini
• Zhu, Chenyun

Project Summary In this project, we built two recommendation algorithms and one predictive model on "The Instacart Online Grocery Shopping Dataset 2017". Video Demo can be found here.

• Recommend new products to customers
• Recommend the bundles of products to customers
• Implement models that predict which products a customer will buy again using transactional data

Project Introduction

What’s your daily routine that keeps you busy the whole day? Go to the gym? Shop your groceries?

Instacart is a same-day grocery delivery service that can save yourself that trip to the market. It will connect you with personal shoppers in your area to shop and deliver groceries from your favorites stores in as little as an hour. Instacart, as a grocery delivery startup, has several main competitors like AmazonFresh and Shipt. In order to acquire more customers and increase customer retention, Instacart aims to make it easy to fill your refrigerator and pantry with your personal favorites when you need them. Also, it tries to provide delightful shopping experience.

For those reasons, we would like to emphasize two problems: recommend customers with their favorite products and predict which products will be in a customer’s next order.