0

I'm trying to get into deep learning. I have this project in mind. Let's say I have a dataset that consists of two columns. Those columns are a list of interests of two matching people. It goes like this: "cars, psychology, politics" "beauty, philosophy, sport". Now, I'm trying to train my model to predict the second person's interest based on the input of the first person's interests.

I have been experimenting with uber's ludwig, which is a deep learning toolbox that you can use without writing code. The examples they provide are great but I was unable to fit them for my desirable model. Maybe you can push me towards the suitable model for this problem?

user1743703
  • 133
  • 2
  • 10
  • 1
    It's not very clear to me what is being predicted here. You have a dataset of many matched pairs of people... for every pair, there is a list of interests for person `A` and a separate list of interests for person `B`. Now, you want to ask the question: given a new person's set of interests, can a model predict the interests that would theoretically match with this new person? If so, I would ask you to considered clustering algorithms that group together interests that tend to match. – Jason K Lai Dec 12 '19 at 19:49
  • Yes, you got it exactly right. Thanks for info. I will look into clustering algorithms – user1743703 Dec 12 '19 at 19:52

0 Answers0