"Using cluster model results and validate"

clememartinez20clememartinez20 Member Posts: 4 Contributor I
edited May 23 in Help
Hi there!
            I have used Kmeans to generate clusters with a training set, based in product's features that have a price. After that I would like to use a Testing set that have no price and i'd like to predict it, selecting the cluster with similar elements. How can i do the last thing?

Regards!
Tagged:

Answers

  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,525   Unicorn
    Hi,
    as it seems to me, you should rather use a regression model instead of a clustering. Clustering only gives you an impression which examples are similiar to each other. Go for regression if you want to predict a continuous value.

    Greetings,
      Sebastian 
  • wesselwessel Member Posts: 537  Guru
    Maybe he is interested in the "goodness" of his clusters.

    Clusters can also be used for classification/regression.
    So you can use classification/regression performance as a measure of cluster "goodness".

    Or did I read something wrong?
Sign In or Register to comment.