0
$\begingroup$

Hi I am finding it hard to find online the best clustering algorithm for clustering people according to answers they gave on 20 question survey. There are four categories which each of these answers can fall into. I want to cluster the respondents according to their category answers, assuming it is multiple choice questions on the survey, with questions only having two options. If person A chose the same or similar answers to person B in a similar category then they should be clustered together.

What is the best algorithm? Sorry, I found it hard to do google searches for this question.

$\endgroup$
  • $\begingroup$ Clustering algorithms are used when we cannot categorize the data. Your problem could be solved with hard-coded rules. There is no need for clustering algorithms. $\endgroup$ – Shubham Panchal May 9 at 6:39
  • $\begingroup$ Thank you Shubham. $\endgroup$ – nealous3 May 11 at 2:21
0
$\begingroup$

Maybe you can

  • assign numeric values to the response of each question (e.x. 0 disagree, 1 for neutral, 2 for agree) if there is a natural order in them,
  • perform one hot encoding for the rest where no such order exist (ex. country of origin)

and then try your luck with a classical algorithm like k-means or hierarchical clustering. There exist a lot of bibliography against that approach but maybe you can create clusters that make sense to your case.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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