Skip to content

Automatic K-Finder #2

Description

@Junhyuck-Woo

Our assignment is to implement k-finder.
I thought that k-finder function gives us k value automatically.

Actually, I could find two methods, the first one is an elbow method and the other is a silhouette method. In the Labs, your description might mean an elbow method.
So, I tried to implement the elbow method.

I have two problems.

  1. If the k value is increased, the error rate is always reduced.
    So, I utilized a differentiation. However, sometimes differential also decreased gradually.
    It made me hard to get a collect one.
    Is there are any solutions?
  2. Based on my result, I hope to check my result.
    Is it correct?

PS.
In the noon, I asked a different one.
I found an error in my algorithms, so I changed it.
This issue is another problem I faced.

Here are images (https://imgur.com/a/caffR9g)
1, 2 show my problem.
3,4,5 for checking my algorithms ( Results)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions