🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.

CLICK HERE TO DOWNLOAD

how to get performance than k-Means Clustering?

goldgold Member Posts: 8 Newbie
Hi everybody
I have a problem. I want to use performance after k-Means Clustering. For this aim I must to use map clustering on labels after clustering and when I run this project I saw an error and I must to changing the number of K, while I am not allowed to change the number of K because I am doing thesis and it not possible for me. Is there any solution for this problem? look at the picture please.
In the second step I thought I might be able to use sample to solve this problem but I saw an error abut sample size. I don't know what is the best sample size in this way? Is this method correct?  look at the picture please( the sample size is 100 in this picture). 
Thank you for your attention.  

Answers

  • ceaperezceaperez Member Posts: 222   Unicorn
    Hi @gold,

    first to all, you have an excellent example model about the classification performance in Rapidminer operator tutorial. please search the Performance Classification (Classification) operator and open their tutorial (Use of performance port in Performance (Classification))
    In response about the error tha you shown in your pictures, the error in  Map Clustering on Labels operator maybe is due to, in order to map the labels and clusters you need to have the same number of elements in your label attribute and clusters.

    I hope this can help you

    Best, 

    Cesar
    lionelderkrikor
Sign In or Register to comment.