Hierarchical Aggregation Approach for Distributed Clustering of Spatial Datasets

Abstract

Abstract-In this paper, we present a new approach of distributed clustering for spatial datasets, based on an innovative and efficient aggregation technique. This distributed approach consists of two phases: 1) local clustering phase, where each node performs a clustering on its local data, 2) aggregation phase, where the local clusters are aggregated to produce global clusters. This approach is characterised by the fact that the local clusters are represented in a simple and efficient way. And The aggregation phase is designed in such a way that the final clusters are compact and accurate while the overall process is efficient in both response time and memory allocation. We evaluated the approach with different datasets and compared it to well-known clustering techniques. The experimental results show that our approach is very promising and outperforms all those algorithms.

Breast tumor localization and segmentation using machine learning techniques: Overview of datasets, findings, and methods

“…Moreover, K-medoids method is a type of K-means algorithm that is more resilient to outliers and noises. K-medoids technique uses an individual point in the cluster to describe it, while K-means uses the cluster's mean point [123], [124].…”

Lung Infection Segmentation for COVID-19 Pneumonia Based on a Cascade Convolutional Network from CT Images

“…Clustering can be outlined as an unsupervised strategy that is aimed at fragmenting the input data (image or signal etc.) into the predefined segments (such as K-means method) or automated recognize parts (such as mean-shift method) based on certain criteria such as differences in the color, magnitude, and location [27][28][29][30]. The fuzzy c-means (FCM) algorithm used in our work is an unsupervised data dividing/splitting strategy.…”

Time‐Frequency Analysis of EEG Signals and GLCM Features for Depth of Anesthesia Monitoring

“…KNN classifies the samples based on the distance between the unknown features and the features of training samples. It considers the labels of K as the most similar neighbors to predict the class of the training samples and considers the label of the class with the greatest number of samples among them [41][42][43][44]. In this study, we consider several distance measures to compute the similarity between the test and training samples.…”