1
$\begingroup$

I've got a data-set with items bought, the time it was bought (I can add the weather of the location at that time of the day). I would like a simple "prediction" model based on time and weather.

Most of the time series predictions I've seen have quantitative data to work with so it could be predicted by using regression, but predicting a categorical feature. Which method/approach do you recommend?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

I don't think That is a time series problem Subash,

You want to predict the item type based on some categorical variables that may be repeat and changed within the same time step.

These categorical variables may include

  1. Hour of day.
  2. Day of Week.
  3. Weather condition

I think it is a classification problem.

If you have any further explanation I would be glad to help.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Thanks for the clarification Amr. My dataset has 4 columns: Date, time, transactionId and Item. What I'm trying to do is basically predict or find the highest probability of you buying a certain item at a certain time of the day. For example, from exploration of the data, I can see that most coffee are bought at 11am. So I want to recommend at around 11am to buy coffee. Hope that makes sense $\endgroup$
    – Subash Poudyal
    Commented Feb 18, 2019 at 14:48
  • $\begingroup$ Yes, it dose. However the critical thing here is how we look at a business problem and map it to a data science problem or a model. I think that you can't deal with this problem as a time series problem, I think you need to think of it as a classification problem and try to extract variable information from these 4 variables that will help the model distinguish between these different item types. On top of my head I can say that, if I am solving this problem I would start with frequency based algorithms.Best of luck $\endgroup$
    – Amr Tarek Elmzayen
    Commented Feb 19, 2019 at 8:58

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.