Labeling clusters

student_computestudent_compute MemberPosts:73Contributor II
edited December 2018 inHelp

Hello friends engineer
I am a beginner . I searched a lot but did not understand. So I ask.
I've done the clustering of the text
Now I want to know what the cluster is about? What about And then I can get information based on that. Labels clusters.
Who knows how to do in rapidshare?
Thankful

Answers

  • rfuentealbarfuentealba Moderator, RapidMiner Certified Analyst, Member, University ProfessorPosts:568Unicorn

    Hi@student_compute:

    Clusteringmeans "grouping". You are letting RapidMiner determine how to create groups from data that looks too random for your eyes. Once you have these groups, you can use the设置角色attribute to change the role of a集群into alabel, and then you can form aDecision Treeto build and interpretation on what makes this cluster different from the others.

    This questioncan help you. Also, here is an article written by@mschmitzonunderstanding clustering with decision trees. It has drawings too.

    @m_keshavarz_comhey, read this! You will need this for the process you are building too.

    Have a nice day!

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, MemberPosts:1,635Unicorn

    You can also look at the centroid output of the cluster model and then sort for the highest values---this will enable you to understand the most prominent attributes in each of your clusters.

    Brian T.
    Lindon Ventures
    Data Science Consulting from Certified RapidMiner Experts
    rfuentealba
  • student_computestudent_compute MemberPosts:73Contributor II

    Hello
    thank you very much dear friend
    I did not understand what exactly these numbers say?
    Can you explain more? What does zero mean?
    How to know the center of the cluster?
    How can I show the shape of the clusters, for example, as a loss to R programming?
    help me
    Thanks a lot

  • student_computestudent_compute MemberPosts:73Contributor II

    Hello
    Sorry someone is helping me
    Thankful

Sign InorRegisterto comment.