search

marcoscastro / kmeans

Implementation of the K-Means clustering algorithm.
MIT License
104 stars 55 forks source link

Convergence #1

Open Raminmoslemi opened 7 years ago

Raminmoslemi commented 7 years ago

Are you sure your algorithm converge? For several run with same data set give different results!! Could you please explain this.

marcoscastro commented 7 years ago

Hi, sorry for delay. Yes, the algorithm converge. The k-means tradicional algorithm has a random inicialization, then for the same dataset give different results.

[]'s

Marcos Castro de Souza

De: Raminmoslemi notifications@github.com Enviado: terça-feira, 20 de dezembro de 2016 22:50 Para: marcoscastro/kmeans Cc: Subscribed Assunto: [marcoscastro/kmeans] Convergence (#1)

Are you sure your algorithm converge? For several run with same data set give different results!! Could you please explain this.

- You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/marcoscastro/kmeans/issues/1, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AFZd8CpimVSXvb4HVrnNPGwWIo84WPsnks5rKHfsgaJpZM4LSdcl.