Shopper Behavior Analysis

This project is about exploring shopper behavior and build a shopping cart recommender using Instacart data.

Introduction

In this project, I want to explore customer shopping behavior by analyzing their pervious purchases and build a shopping cart recommender that could give them a tailored shopping experience to drive engagement.

I want to perform a Shopper Behavior Analysis by analyzing shoppers purchase history and see if I could find patterns such as what kind of items customers like to purchase in different time of day/week, how often do people reorder the same product they purchased before and etc. And build a recommender system using customer's reordered purchases and cosine similarity.

For complete write-up of this project, please go to my post on medium.

Project Design

1. Perform Data Cleaning, merging different tables together and do Exploratory Data Analysis (EDA).
2. Analyze customer behavior, generate insights and build data visualizations for storytelling.
3. Define a metric for model evaluation.
4. Lastly, create a shopping cart recommender system to recommend items to customers based on their purchase history.

Getting Started

• notebooks - My code for exploratory data analysis, data visualizations, recommender system and metirc calucation are in this folder.
• presentation - This is the deck that I used to present in Metis. For references, please refer to the reference page in my deck.