Is there any way to run any clustering model on a small dataset with 290 text records (minimum character size is 100)?

  • 1
    $\begingroup$ Yes, but the accuracy or correctness of clustering will depend on how you conduct the feature extraction. And probably lower the feature space, better will it be for the clustering algorithm. $\endgroup$ Nov 26, 2021 at 4:39

1 Answer 1


Certainly. I demonstrated clustering techniques using 52 records (playing cards and their features). Unlike classification algorithms, clustering will work with the data available.

The question to ask yourself is whether your data has a sufficient number of features that are enable records to be both clusterable and separable.

  • $\begingroup$ Thanks, Anandologist. What clustering method did you use? $\endgroup$
    – SaNa
    Nov 28, 2021 at 23:01
  • $\begingroup$ In the case of the playing cards, I was demonstrating how different clustering algorithms would perform. I did that using k-means, agglomerative, and DBSCAN. $\endgroup$ Nov 29, 2021 at 13:14

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.