On a Framework for Federated Cluster Analysis

Federated learning is becoming increasingly popular to enable automated in distributed networks of autonomous partners without sharing raw data. Many works focus on supervised learning, while the area federated unsupervised similar clustering, still less explored. In this paper, we introduce a clustering framework that solves three challenges: determine number global clusters dataset, obtain partition data via fuzzy c-means algorithm, and validate through Davies–Bouldin index. The complete evaluated numerical experiments artificial real-world datasets. observed results are promising, as most cases framework’s consistent with its nonfederated equivalent. Moreover, embed an alternative formulation into our observe more reliable case noni.i.d., performance par i.i.d. case.