kawong14 / amazon_beauty_products

Fully Connected Neural Network and XGBoost for Predicting Beauty Products

Amazon Beauty Products

Key Problems / Issues:

As a beauty retailer on Amazon, you want to determine if there is an optimal way you can recommend products to customers

Neural Network:

• The purpose of this neural network is to make predictions to see if customers would like a product
• The three variables I am looking at are (User ID, Product ID, and Rating)
• By creating a neural network on users’ history of the products and how they rated it the model would recommend products
  • Inputs:
    • User ID (reviewerID)
    • Product ID (asin = amazon standard identification number)
  • Output: Rating (overall)

How It Works:

For Example: We want to predict products Customer-B would likely want to purchase

• Customer-A purchases Product 1 and rates it 5/5 stars with 5 being the highest value
• Customer-A then purchases Product 2 and rates it 5/5 stars
• Customer-B purchases Product 1 and rates it 5/5 stars Therefore the model would recommend that Customer-B would like Product 2

Analyses:

Option 1: Fully Connected Neural Network
Option 2: Gradient Boosted Decision Trees-XGBoost

Summary:

The results from Option1 had an RMSE of 1.32, and the Option2 has an RMSE of 0.33. The XGBoost model can more accurately predict the rating of a product than the neural network.