Amazon-ML-Engineer-Hiring-Challenge

Problem Statement

We all know that when we visit an e-commerce or TV series website or even YouTube we see a separate suggestion box, where in they show some content which you might like. These are mainly based on the content that you have consumed on their website previously. These are called as Recommendation engine.

Now consider you have been running a start up since last one year and now you have been able to gather some customer data and you want to build a recommendation engine. Based on certain features you have to cluster the customers into two different groups so that you can recommend the correct products based on the customer’s cluster.

Challenge

Your work is to build a predictive model to predict the category of the customer. You have to predict the column : “customer_category”

You need to upload a presentation proposal addressing the following questions:

• What are the pros & cons of recommendation by this approach?
• Propose an architecture that will work more efficiently when building a recommendation engine for an e-commerce platform