China
Africa
Explore chapters and articles related to this topic
Supervised and Semi-Supervised Identification of Users and Activities from Wearable Device Brain Signal Recordings
Published in B.K. Tripathy, J. Anuradha, Internet of Things (IoT), 2017
Glavin Wiechert, Matt Triff, Zhixing Liu, Zhicheng Yin, Shuai Zhao, Ziyun Zhong, Runxing Zhou, Pawan Lingras
This section describes a variation of the evolutionary rough K-means approach proposed by Lingras [7]. The proposal replaces K-means with K-medoids. A medoid is the most centrally located object in a given cluster. For k clusters, we will have K-medoids. A genetic algorithm can be used to search for the most appropriate K-medoids. The genome will contain k genes, each corresponding to a medoid. This reduces the size of a genome from km by a factor of m to k. The smaller genomes will reduce the space requirements and facilitate faster convergence. The genes in the rough K-means algorithm were continuous real variables with no restriction on their values. The values of genes for the medoids will be discrete and limited to the number of objects in the dataset. If we number the objects from 1,…,n, then each gene can take an integer value in the range 1,…,n. This restriction on the values of genes will further reduce the search space allowing for even faster convergence.
Unsupervised Learning
Published in Peter Wlodarczak, Machine Learning and its Applications, 2019
In centroid-based clustering, the data points are grouped around a centroid, the mean of all the objects in a cluster. A centroid usually does not correspond to an actual data point. By contrast, a medoid must be a data point. A medoid is the most representative data point in a cluster. It is used when there is no meaningful centroid, e.g., when the data is categorical and not continuous. The prototype in prototype-based clustering defines the cluster. Each data point is closer or more similar to the prototype than a prototype in any other cluster. The prototype is often the centroid. The prototype can be regarded as the most central point and the clusters tend to be globular. If the clusters have no centroid, the prototype can be the medoid.
A Survey of Partitional and Hierarchical Clustering Algorithms
Published in Charu C. Aggarwal, Chandan K. Reddy, Data Clustering, 2018
Chandan K. Reddy, Bhanukiran Vinzamuri
K-medoids is a clustering algorithm which is more resilient to outliers compared to K-means [38]. Similar to K-means, the goal of K-medoids is to find a clustering solution that minimizes a predefined objective function. The K-medoids algorithm chooses the actual data points as the prototypes and is more robust to noise and outliers in the data. The K-medoids algorithm aims to minimize the absolute error criterion rather than the SSE. Similar to the K-means clustering algorithm, the K-medoids algorithm proceeds iteratively until each representative object is actually the medoid of the cluster. The basic K-medoids clustering algorithm is given in Algorithm 14.
Analysis of Price Based Demand Response Program Using Load Clustering Approach
Published in IETE Journal of Research, 2023
Jayesh Priolkar, E. S. Sreeraj
The medoids are selected randomly from the dataset to form the K clusters and other data points are placed near the medoids in the cluster. All the data objects of the cluster are processed to find the new medoids in a repeated manner which represents the new cluster in a better way. Every point is tested as a potential medoid within the cluster by finding out whether the sum within the cluster distances gets smaller using that point as a medoid. Based on the new medoids all the objects get bound to the clusters, for each iteration medoids change the location step by step and the process is continued till no medoids move. Due to pairwise distance measures in the case of K-medoids, the overlapping reduces as compared to K-means. The output of the algorithm gives medoids which are actual data points in the data set.
An XNOR-ResNet and spatial pyramid pooling-based YOLO v3-tiny algorithm for Monkeypox and similar skin disease detection
Published in The Imaging Science Journal, 2023
The series of YOLO algorithms determine anchor boxes using k-means clustering. The anchor boxes measure the quality of detection of objects under consideration. The object detection algorithm can detect multiple objects based on the recognized annotations however, for precise prediction anchor boxes plays a vital role. In this work, as a step to obtain better values of anchor boxes, we have used the k-medoids clustering [29] mechanism. The advantage of using k-medoids clustering is that it minimizes the sum of variance between the points of labelled bounding boxes and the centre of those bounding boxes. Whereas, the k-means clustering contributes to minimizing the total squared between the bounding boxes and their centre. Moreover, the k-medoids clustering is more robust to outliers.
Cluster analysis and multi-level modeling for evaluating the impact of rain on aggressive lane-changing characteristics utilizing naturalistic driving data
Published in Journal of Transportation Safety & Security, 2022
Anik Das, Md Nasim Khan, Mohamed M. Ahmed, Shaun S. Wulff
In contrast to the K-means clustering, the K-medoids method selects representative observations as centers, which are called medoids and observations cluster around those medoids. The K-medoids clustering can be used with random dissimilarity measures. The error function of this method is calculated by the summation of the dissimilarities between each observation and its respective medoid (Mohammadnazar, Arvin, & Khattak, 2021; Park & Jun, 2009), as shown in Equation 4. where denotes cluster and represents the medoid of the cluster Note that several algorithms were proposed for K-medoids clustering; however, this study considered one of the widely used K-medoids algorithms called Partitioning Around Medoids (PAM) (Kaufman & Rousseeuw, 2009).