Suppose I buy products: [1,2,3,4]

Another customer X bought: [2,3]

Most probably customer X next purchase will be: 4

Sequence is very important in my problem

I tried association analysis using R, but it don't take under consideration the sequence

Please advise what algorithms I need to solve this?

Do I need to First do Clustering to find similar customers?


1 Answer 1


For sequence problems, I generally used Recurrent Neural Network. It has a property to learn its values, which is based upon the previous state and current input. Since in your case sequence is important, you can use RNN. You can also try LSTM(type of RNN) cell.

And Additional Note: For product suggestion(where sequence is not important) you can also use Apriori algorithm. This algorithm try to build association between the products of a single order.

  • $\begingroup$ Could you please share with me useful reference or links with examples for product suggestion using RNN $\endgroup$
    – sara
    Jul 9, 2019 at 4:43
  • $\begingroup$ For series problem you can refer link "machinelearningmastery.com/…", commonly used term is time series problem. In your case instead of time, you feed your product list(order is important as you said), in a order and it will predict the next product. $\endgroup$ Jul 9, 2019 at 5:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.