How to Categorize Uncategorized Data

  • Stephan Calderon

    SSC Veteran

    Points: 267

    Comments posted to this topic are about the item <A HREF="/articles/Cluster+Analysis/124259/">How to Categorize Uncategorized Data</A>

    I'm very sorry that the CA_Example_Instructions.zip was not uploaded. Check attached or the link below:

    https://www.dropbox.com/s/gpet8hrk7ntxnnq/CA_Example_Instructions.zip?dl=0

    Also there are some formatting issues when I copied the article from Word.

    In step four of Silhouette formula, it should say

    Silhouette Score = (Y-X)/C

    Sincerely,

    Stephan Calderon

  • aspas

    SSC Journeyman

    Points: 87

    Hi!

    Nice article! We are doing this with R but I have always wondered how to transfer it to T-SQL.

    I can't find the link to the CA_Example_Instructions.zip file mentioned in the article, can you please post it?

    Thanks a lot!

    Xavi

  • manzoor926

    SSC Rookie

    Points: 29

    where is zip file and sql file .

  • chris.puncher

    SSC-Addicted

    Points: 433

    Also can't find the SQL.

    In the absence of this I'm trying to understand the silhouetting explanation. How is the result in the range -1 to 1? Are the distances limited (is the scale on the cluster diagram a percentage)? Even so if you take the greater of the averages X and Y, how can you ever end up with a negative value?

  • akljfhnlaflkj

    SSC Guru

    Points: 76202

    It's all a bit confusing to me. I'll reread.

  • Stephan Calderon

    SSC Veteran

    Points: 267

    I updated the original post of this forum to include CA_Example_Instructions.zip

  • aspas

    SSC Journeyman

    Points: 87

    I can't see the link anyway, where did you place the link or the file?

    Thank you!

  • Stephan Calderon

    SSC Veteran

    Points: 267

    There was an issue in formatting. I updated the topic to show the correct formula for Silhouette.

    Sorry for the confusion. If you have any more questions, don't hesitate to ask

    -Stephan Calderon

  • Stephan Calderon

    SSC Veteran

    Points: 267

    It's at the top of this forum. It's the first post(call Topic I think).

    Sorry, I'm very new to all of this.

  • aspas

    SSC Journeyman

    Points: 87

    Thanks a lot!!!

  • susan.wong

    SSC Enthusiast

    Points: 107

    Article says yellow cluster to be junior high girls, should it be blue?

  • Stephan Calderon

    SSC Veteran

    Points: 267

    Thank you Susan for the question.

    It's a simple example so it doesn't really matter which color is which. Girls are yellow because that's how my report returned it.

    And you'll notice the legend actually labels the yellow as the girls group.

  • mbellil

    Valued Member

    Points: 65

    Bi

    some one answer to the question of SSC ROOKIE??

    In calculating silhouetting ...i n step 3. How do i know hkw many cluster i do have ??? Did u use the ssrs graph to see how many cluster i the data?

    Thank u

Viewing 13 posts - 1 through 13 (of 13 total)

You must be logged in to reply to this topic. Login to reply