资 源 简 介
Distributed K-means Algorithm
K-means clustering is a well-known and well-studied exploratory data analysis technique.
The standard version assumes that all data are available at a single location. The system deployed in this project implements a dynamic distributed k-means algorithm which has been described in the section 5 of the paper “Approximate Distributed K-Means Clustering over a Peer-to-Peer Network”.
It takes a completely decentralized approach, where peers (nodes) only synchronize with their immediate neighbors. The result of this distributed k-means algorithm has been tested
to achieve very good accuracy (more than 97%, mentioned in the section 6.5.1 of the paper)
compared to centralized k-means clustering.
Goal
Aiming to realize a distributed kmeans clustering algorithm which allows dynamic change of nodes.
Still in the process of changing from static clustering to dynamic one. There are three steps:
1) allowing the
