Share via

different results in clustering

Sanniddha Chakrabarti 36 Reputation points
2020-12-23T17:29:30.72+00:00

One of the disadvantage of clustering is the result can be different, because it randomly selects the initial mean point. It should impact on the time to take complete the clustering, but how it effects on the result?

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
3,332 questions
Community Center Not monitored
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Ramr-msft 17,826 Reputation points
    2020-12-24T14:58:06.333+00:00

    @Sanniddha Chakrabarti Thanks, You can use silhouette clustering to find out and validate the optimal number of cluster .
    Reference – sklearn , wiki

    running a cluster elbow analysis. This is a method that worked well for me for example 1M rows and 5 features (computation time of about 10-15 seconds)

    https://www.geeksforgeeks.org/elbow-method-for-optimal-value-of-k-in-kmeans/

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.