K-medoids Clustering Research Articles

Optimization of Markov Weighted Fuzzy Time Series Forecasting Using Genetic Algorithm (GA) and Particle Swarm Optimization (PSO)

The Markov Weighted Fuzzy Time Series (MWFTS) is a method for making predictions based on developing a fuzzy time series (FTS) algorithm. The MWTS has overcome certain limitations of FTS, such as repetition of fuzzy logic relationships and weight considerations of fuzzy logic relationships. The main challenge of the MWFTS method is the absence of standardized rules for determining partition intervals. This study compares the MWFTS model to the partition methods Genetic Algorithm-Fuzzy K-Medoids clustering (GA-FKM) and Fuzzy K-Medoids clustering-Particle Swarm Optimization (FKM-PSO) to solve the problem of determining the partition interval and develop an algorithm. Optimal partition optimization. The GA optimization algorithm’s performance on GA-FKM depends on optimizing the clustering of FKM to obtain the most significant partition interval. Implementing the PSO optimization algorithm on FKM-PSO involves maximizing the interval length following the FKM procedure. The proposed method was applied to Anand Vihar, India’s air quality data. The MWFTS method combined with the GA-FKM partitioning method reduced the mean absolute square error (MAPE) from 17.440 to 16.85%. While the results of forecasting using the MWFTS method in conjunction with the FKM-PSO partition method were able to reduce the MAPE percentage from 9.78% to 7.58%, the MAPE percentage was still 9.78%. Initially, the root mean square error (RMSE) score for the GA-FKM partitioning technique was 48,179 to 47,01. After applying the FKM-PSO method, the initial RMSE score of 30,638 was reduced to 24,863. Doi: 10.28991/ESJ-2022-06-06-010 Full Text: PDF

Three-dimensional inversion of knot defects recognition in timber cutting

The comprehensive utilization of wood is the main goal of log cutting, but knot defects increase the difficulty of rationally optimizing cutting. Due to the lack of real shape data of knot defects in logs, it is difficult for detection methods to establish a correlation between signal and defect morphology. An image-processing method is proposed for knot inversion based on distance regularized level set segmentation (DRLSE) and spatial vertex clustering, and with the inversion of the defects existing relative board position in the log, an inversion model of the knot defect is established. First, the defect edges of the top and bottom images of the boards are extracted by DRLSE and ellipse fitting, and the major axes of the ellipses made coplanar by angle correction; second, the coordinate points of the top and bottom ellipse edges are extracted to form a spatial straight line; third, to solve the intersection dispersion of spatial straight lines and the major axis plane, K-medoids clustering is used to locate the vertex. Finally, with the vertex and the large ellipse, a 3D cone model is constructed which can be used to invert the shape of knots in the board. The experiment was conducted on ten defective larch boards, and the experimental results showed that this method can accurately invert the shapes of defects in solid wood boards with the advantages of low cost and easy operation.

A Hybrid Method for Traffic State Classification Using K-Medoids Clustering and Self-Tuning Spectral Clustering

As an important part of intelligent transportation systems, traffic state classification plays a vital role for traffic managers when formulating measures to alleviate traffic congestion. The proliferation of traffic data brings new opportunities for traffic state classification. In this paper, we propose a hybrid new traffic state classification method based on unsupervised clustering. Firstly, the k-medoids clustering algorithm is used to cluster the daily traffic speed data from multiple detection points in the selected area, and then the cluster-center detection points of the cluster with congestion are selected for further analysis. Then, the self-tuning spectral clustering algorithm is used to cluster the speed, flow, and occupancy data of the target detection point to obtain the traffic state classification results. Finally, several state-of-the-art methods are introduced for comparison, and the results show that performance of the proposed method are superior to comparable methods.

Clinical complexity and impact of the ABC (Atrial fibrillation Better Care) pathway in patients with atrial fibrillation: a report from the ESC-EHRA EURObservational Research Programme in AF General Long-Term Registry

BackgroundClinical complexity is increasingly prevalent among patients with atrial fibrillation (AF). The ‘Atrial fibrillation Better Care’ (ABC) pathway approach has been proposed to streamline a more holistic and integrated approach to AF care; however, there are limited data on its usefulness among clinically complex patients. We aim to determine the impact of ABC pathway in a contemporary cohort of clinically complex AF patients.MethodsFrom the ESC-EHRA EORP-AF General Long-Term Registry, we analysed clinically complex AF patients, defined as the presence of frailty, multimorbidity and/or polypharmacy. A K-medoids cluster analysis was performed to identify different groups of clinical complexity. The impact of an ABC-adherent approach on major outcomes was analysed through Cox-regression analyses and delay of event (DoE) analyses.ResultsAmong 9966 AF patients included, 8289 (83.1%) were clinically complex. Adherence to the ABC pathway in the clinically complex group reduced the risk of all-cause death (adjusted HR [aHR]: 0.72, 95%CI 0.58–0.91), major adverse cardiovascular events (MACEs; aHR: 0.68, 95%CI 0.52–0.87) and composite outcome (aHR: 0.70, 95%CI: 0.58–0.85). Adherence to the ABC pathway was associated with a significant reduction in the risk of death (aHR: 0.74, 95%CI 0.56–0.98) and composite outcome (aHR: 0.76, 95%CI 0.60–0.96) also in the high-complexity cluster; similar trends were observed for MACEs. In DoE analyses, an ABC-adherent approach resulted in significant gains in event-free survival for all the outcomes investigated in clinically complex patients. Based on absolute risk reduction at 1 year of follow-up, the number needed to treat for ABC pathway adherence was 24 for all-cause death, 31 for MACEs and 20 for the composite outcome.ConclusionsAn ABC-adherent approach reduces the risk of major outcomes in clinically complex AF patients. Ensuring adherence to the ABC pathway is essential to improve clinical outcomes among clinically complex AF patients.

Clustering Strategies for Defining Archetypes to Support Integrated Simulations of Environmental Impacts

Life cycle assessment (LCA)’s inherent data-intensiveness hampers application to neighbourhood environmental assessments, particularly for built stock modelling. Data collection can be reduced to a manageable amount by grouping a large number of buildings into a limited set of aggregates with similar characteristics and defining exemplars (archetypes) that represent each group. LCAs would be performed for the archetypes only and their results extended to the represented buildings. As archetype definition is seldom detailed in the literature, this paper tests, and details different procedures that could enable neighbourhood LCAs. K-medoids and CLARA partition algorithms, as well as agglomerative hierarchical clustering techniques, were applied to group over 300 buildings into a limited number of clusters. A building representative of each cluster was identified to proceed to bottom-up LCA. K-medoids clustering stands out for the quality of clusters and their representatives. Restraining the maximum number of clusters to keep subsequent LCA work manageable imposes some quality loss yet allows for achieving satisfactory division results. Regardless of the clustering technique used, data was the best divided the larger the number of clusters used, for the various factors in the database depicting the studied area resulted in several possible data combinations. Although detailed representation is desirable in LCA modelling, limiting the number of variables facilitates data pre-treatment and an optimal balance should be pursued in future studies.

Comparison of Hierarchical Clustering, K-Means, K-Medoids, and Fuzzy C-Means Methods in Grouping Provinces in Indonesia according to the Special Index for Handling Stunting

Stunting has been widely known as the highest case of malnutrition suffered by toddlers in the world and has a bad impact on children's future. In 2018, Indonesia was ranked the 31st highest stunting in the world and ranked 4th in Southeast Asia. About 30.8% (roughly 3 out of 10) of children under 5 years suffer from stunting in Indonesia. To support the government policy making in handling stunting, it is undoubtedly necessary to classify the levels of stunting handling in regions in Indonesia. In this work, the hierarchical agglomerative and non-hierarchical clustering is compared and evaluated to perform clustering on stunting data. The agglomerative hierarchical cluster uses Single Linkage, Average Linkage, Complete Linkage, and Ward Method, while the non-hierarchical cluster uses K-Means, K-Medoids (PAM) Clustering, and Fuzzy C-Means. This study uses data from 12 IKPS indicators in 34 provinces in Indonesia in 2018. Based on the results of the evaluation using the Connectivity Coefficient, Dunn Index, Silhouette Coefficient, Davies Bouldin Index, Xie & Beni Index, and Calinski-Harabasz Index, the results show that the Average Linkage is the best cluster method with the optimal number of clusters is four clusters. The first cluster is a cluster with a good level of stunting management which consists of 28 provinces. The second cluster consists of only one province, DI Yogyakarta with a very good level of stunting handling. The third cluster consists of four provinces with poor stunting handling rates. Finally, the last cluster consisting of one province, Papua, has a very poor level of stunting handling.

A Multi-step ahead photovoltaic power forecasting model based on TimeGAN, Soft DTW-based K-medoids clustering, and a CNN-GRU hybrid neural network

Accurate photovoltaic (PV) power generation forecasting is very important for making economic and reliable power dispatching plans. This study proposes a multi-step ahead PV power forecasting (PPF) model, which combines time-series generative adversarial networks (TimeGAN), soft dynamic time warping (DTW)-based K-medoids clustering algorithms, and a hybrid neural network model computed by a convolutional neural network (CNN) and gated recurrent units (GRU). First, a new PV power data augmentation method based on TimeGAN is proposed to realize the expansion of the historical PV power data. Second, the combination of soft-DTW and K-medoids clustering is developed to extract and distinguish the commonalities and differences between time series characteristics of PV power. Finally, a hybrid neural network model is developed to integrate CNN and GRU into a unified framework for an accurate PPF under different weather conditions. The performance of the proposed model was tested using real datasets of two PV power stations with an installed capacity of 50MW. The comparative analysis of the proposed model with five reference models confirms its superiority in terms of forecasting accuracy. For the sunny days and cloudy days, the range of root mean square error (RMSE) values are produced by the proposed model distributed in [0.927–2.523] MW and [1.781–6.173] MW respectively with respect to the multi-step ahead forecasting from 15-min up to 24-hour. The experimental results of the hybrid model proposed in this study indicate that data augmentation-based clustering method is very helpful for improving the accuracy of PV power forecasting.

Optimal scan planning for surveying large sites with static and mobile mapping systems

Since the last two decades, the use of laser scanners for generating accurate and dense 3D models has been rapidly growing in multiple disciplines. The reliance on human-expertise to perform an efficient scanning in terms of completeness and quality encouraged the researchers to develop strategies for carrying out an optimized and automated scan planning. Nevertheless, due to the predominant use of static terrestrial laser scanners (TLS), the most of developed methods have been focused on scan optimization by fixing standpoints on basis of static scanning. The increasing use of portable mobile laser scanning systems (MLS) enables faster non-stop acquisition which demands the planning of optimal scan trajectories. Therefore, a novel method addressing the absence of dynamic scan planning is proposed considering specific MLS constraints such as maximum acquisition time or closed-loops requirement. First, an initial analysis is carried out to determinate key-positions to reach during data acquisition. From these positions a navigable graph is generated to compute routes satisfying specific MLS constraints by a three-step process. This starts by estimating the number of routes necessary to subsequently carry out a coarse graph partition based on Kmedoids clustering. Next, a balancing algorithm was implemented to compute a balanced graph partition by node exchanging. Finally, partitions are extended by adding key nodes from their adjacent ones in order to provide a desirable overlapping between scans. The method was tested by simulating three laser scanner configurations in four indoor and outdoor real case studies. The acquisition quality of the computed scan planning was evaluated in terms of 3D completeness and point cloud density with the simulator Helios++.

Analysis of Data classification accuracy using ANFIS algorithm modification with K-Medoids clustering

The ANFIS algorithm is a technique in data mining that can be used for the data classification process. The ANFIS algorithm still has weaknesses, especially in determining the initial parameters for the network training process. Thus, an additional algorithm or modification is needed for the determination of these parameters. In this study, a clustering method will be proposed, namely K-Medoids Clustering as an additional method to the ANFIS algorithm. Basically, the ANFIS algorithm uses the FCM (Fuzzy C-Means Clustering) algorithm for the initial initialization of network parameters. The use of this method can cause local minima problems, where the clustering results obtained are not optimal because the pseudo-partition matrix generation process is carried out randomly. The matrix value will determine the initial parameter value in the ANFIS algorithm used in the first layer. Based on the research that has been done, it can be concluded that the accuracy of data classification using the ANFIS algorithm which has been modified with the proposed method provides a fairly good influence in conducting training and classification testing. The increase that occurs in the proposed method is 0.73% for the average training accuracy and an increase of 0.66% for the average testing accuracy.

Constructing a health indicator for bearing degradation assessment via an unsupervised and enhanced stacked autoencoder

To eliminate the noise from the extracted health indicator (HI) and reduce dependence on manual experience including knowledge for selecting the appropriate time–frequency domain indicators and conversions for data pre-processing, multiple-step processing, and multiple model combinations. An enhanced stacked autoencoder (ESAE) based on an exponent weight moving average (EWMA) in a deep learning model is proposed. This ESAE uses the vibration amplitude of an unlabeled original time-domain signal to construct the HI directly. To demonstrate our proposed method is better than other models, including the stacked autoencoder (SAE), stacked denoising autoencoder (SDAE), root mean square (RMS), kurtosis, K-medoids clustering, and self-organizing map (SOM) neural network models. The proposed model is simulated for multiple bearings in two case studies. The experiment result shows that the extracted HI curve is smoother than mentioned-above other models reduce the wrong judgment of bearing health caused by noise effectively. Moreover, high monotonicity is lay a good foundation for following reaming useful life prediction. An index, named Mon, is used to assess the monotonicity performance of all models, the experiment result also shows that the extracted HI by our proposed is superior to other models.

Medoid-based Clustering pada Kecamatan di Kabupaten Lebak dan Pandeglang Provinsi Banten Berdasarkan Trilogi Ketahanan Pangan

Lebak and Pandeglang Regions in Banten Province have a high stunting prevalence of children under 5 years old and have the lowest value of food security index among regions in Banten Province. Cluster analysis to group districts in Lebak and Padeglang Regions is indispensable to characterize the district members in those two regions. The variables applied to calculate distance between districts in a simple k-medoid clustering were trilogy of food security namely the availability, access, and utility of the food from Bureau of Statistics of Lebak and Pandeglang Regions 2019 data. The distances were varied among Euclidean, squared Euclidean, and Manhattan distances. The clustering result was then validated via consensus clustering and internal validation. The suitable number of clusters was four defined as the available and access cluster (cluster 1), the access cluster (cluster 2), the vulnerable cluster (cluster 3), and the available cluster (cluster 4). The cluster 3 as the vulnerable cluster should be focused on because it consists of 38% from overall districts in Lebak and Banten Regions.

Implementation of K-Medoids Clustering Method for Indihome Service Package Market Segmentation

IndiHome (Indonesia Digital Home) is a leading digital fibre optic service product consisting of fibre optic internet services, landline telephones, and interactive TV services. Although the coverage of Indihome products is extensive in the city of Medan, in marketing, Indihome products have not reached the planned target. Based on data from Indihome service package users that have been received, Indihome product users only numbered 6419 customers in all STOs in Medan City. At the same time, the target was planned by PT. Telkom Access Medan, namely Marketing Indihome products, must reach 5,000 customers per month in all STOs in Medan City. Indihome product marketing is an obstacle for PT. Telkom Access Medan, because the Indihome product is a new product, the people of Medan City do not fully know what Indihome is and what facilities they get from using the Indihome service package. Therefore PT. Telkom Access Medan needs to make a plan to make a marketing strategy. The first step that needs to be done is to segment the market for the Indihome service package. This study aimed to determine the application of Data Mining using the K-Medoids Clustering method in the Indihome service package market segmentation at PT. Telkom Access Medan. With this research, it is hoped that it can provide a reference for the results of the decision so that it can help related parties to make it easier to classify the market segmentation of the Indihome service package at PT. Telkom Access Medan. Because the value of S > 0, then the calculation is stopped and ends in the 3rd iteration. Indihome service package data processing uses the k-medoids clustering method in the form of potential, potential, and not potential STO (Sentral Telephone Automated) cluster members.

Hybrid cluster made content forwarding and light weight sign encryption IoT based Named Data Network using African Buffalo optimization algorithm

Information Centric Network (ICN) is a newer technology in handling web content distribution that has recently emerged in order to tackle the risk of data security. For handling content distribution, ICN provides data security via a name-based approach. Named Data Networking (NDN) is an ongoing ICN realisation that was incorporated recently. Named Data Networking (NDN) has recently grown in popularity and significance as a new internet design that solves certain limitations in traditional internet communications. NDN is perfectly adapted for the Internet of Things (IoT), which is today dominated by huge, and emerging applications. In this work, we propose an IoT enabled hybrid cluster-based routing protocol with mitigation of content poisoning attack for information-centric Wireless Sensor Network (WSN)-NDN. In this method, hybrid K-medoids clustering is used with African Buffalo Optimization Algorithm (ABOA), which is to find an optimal shortest path between the cluster heads, and light weight encryption. It is developed by using Hyperelliptic Curve Cryptography (HCC) to mitigate content poisoning. Our proposed system has effective data security as it has encrypted data in the cluster head. The smart health care monitoring system has been used for our proposed method. The proposed method has been subjected to extensive analysis by comparing with other existing methods that should improve performance justified in terms of several metrics by introducing the malicious nodes (10%, 20%, and 30%).

Selecting Representative Samples From Complex Biological Datasets Using K-Medoids Clustering.

Rapid growth of single-cell sequencing techniques enables researchers to investigate almost millions of cells with diverse properties in a single experiment. Meanwhile, it also presents great challenges for selecting representative samples from massive single-cell populations for further experimental characterization, which requires a robust and compact sampling with balancing diverse properties of different priority levels. The conventional sampling methods fail to generate representative and generalizable subsets from a massive single-cell population or more complicated ensembles. Here, we present a toolkit called Cookie which can efficiently select out the most representative samples from a massive single-cell population with diverse properties. This method quantifies the relationships/similarities among samples using their Manhattan distances by vectorizing all given properties and then determines an appropriate sample size by evaluating the coverage of key properties from multiple candidate sizes, following by a k-medoids clustering to group samples into several clusters and selects centers from each cluster as the most representatives. Comparison of Cookie with conventional sampling methods using a single-cell atlas dataset, epidemiology surveillance data, and a simulated dataset shows the high efficacy, efficiency, and flexibly of Cookie. The Cookie toolkit is implemented in R and is freely available at https://wilsonimmunologylab.github.io/Cookie/.

ANALISIS K-MEDOIDS CLUSTERING PADA EPISENTRUM GEMPA BUMI DI PROVINSI SUMATERA BARAT DAN SEKITARNYA

Sumatra is located along the confluence line of the Indo-AustralianPlate and the Eurasian plate, which has high seismic activity. Theisland of Sumatra is prone to earthquakes which pose a considerabledanger, and one of the provinces on this island that is prone toearthquakes is the Province of West Sumatra. So it is necessary tostudy each earthquake characteristic in the Province of West Sumatraand its surroundings using K-Medoids Clustering analysis.Earthquake data used is tectonic earthquake data from December 1970to September 2021 with attributes: latitude, longitude, and magnitude.In this study, several clusters were tried, namely K=2, K=3, K=4, andK=5 using the Euclidean Distance. The most optimum cluster is K=2of 0.448072 using the Silhouette coefficient validation test, theepicenter was at 99.290E and -1.050S with a magnitude of 5.38 on theSR for the first group and the second group produced an epicenter at101.150E and -3.440S with a magnitude of 5.36 SR.

Analyzing the multiscale patterns of jobs-housing balance and employment self-containment by different income groups using LEHD data: A case study in Cincinnati metropolitan area

Achieving a balanced jobs-housing relationship has been treated as a solution to relieve traffic congestion and air pollution. Jobs-housing ratio and employment self-containment (ESC) are applied to quantify different levels of jobs-housing balance. Though appropriated methods have been analyzed to explore jobs-housing relationship at multiple fixed scales, scale dependency problem remains unsolved and resulted in discord about jobs-housing relationship. Furthermore, the differences of the jobs-housing balance and ESC by various income groups received little attention. This paper proposed a K-medoids clustering method to explore the aggregate multi-scale patterns of jobs-housing ratio and ESC by higher, medium, and lower-income groups using 2016 longitudinal employment household dynamics (LEHD) data. To alleviate the scale effect, individual workers and jobs were simulated by Monte Carlo approach according to LEHD data provided within each block. Afterwards, we analyzed the jobs-housing relationship within multi-scale units by different aggregated groups. We found jobs-housing balance had a positive effect on ESC, while the effect varied by different income groups. Jobs-housing balance affects ESC more for lower-income group when the searching radius is <4 km, while beyond 4 km, jobs-housing balance has more effect on ESC of higher-income group. ESC of medium-income group is least affected. Employing the K-medoids clustering method, we classified the four patterns of jobs-housing and the two patterns of ESC for each income group. We further identified jobs-housing matched and mismatched areas through eight joint patterns of jobs-housing balance and ESC. Our work contributes to understanding of jobs-housing balance and commuting by scale variations and income.

Research on Segmenting E-Commerce Customer through an Improved K-Medoids Clustering Algorithm.

In view of the shortcomings of traditional clustering algorithms in feature selection and clustering effect, an improved Recency, Frequency, and Money (RFM) model is introduced, and an improved K-medoids algorithm is proposed. Above model and algorithm are employed to segment customers of e-commerce. First, traditional RFM model is improved by adding two features of customer consumption behavior. Second, in order to overcome the defect of setting K value artificially in traditional K-medoids algorithm, the Calinski–Harabasz (CH) index is introduced to determine the optimal number of clustering. Meanwhile, K-medoids algorithm is optimized by changing the selection of centroids to avoid the influence of noise and isolated points. Finally, empirical research is done using a dataset from an e-commerce platform. The results show that our improved K-medoids algorithm can improve the efficiency and accuracy of e-commerce customer segmentation.

Selecting representative samples from complex biological datasets using k-medoids clustering

Selecting representative samples from complex biological datasets using k-medoids clustering

ROBUST CLUSTERING OF COVID-19 PANDEMIC WORLDWIDE

COVID-19 pandemic is described as the most challenging crisis that humans have faced since World War II. From December 2019 until August 2021 based on the dataset provided by WHO, globally 219 countries in the world are affected by this virus. There are 205.338.159 cases cumulative total and 4.333.094 death cumulative total caused by this virus. In this paper, the data of 219 countries are analyzed using a robust clustering method namely K-Medoids cluster analysis. Based on the result, 219 countries in the world can be divided into five clusters based on four COVID-19-related variables, i.e. the number of cases cumulative total, death cumulative total, positive cases per capita, and case fatality rate. The distribution of the countries in five clusters was as follows; the first cluster contained 48 countries, the second cluster contained 3 countries, the third and fourth clusters contained 16 and 89 countries respectively, and the last cluster contained 63 countries. The largest cluster is the fourth one, containing countries that form a cluster with a centroid below the world average, and the smallest cluster is the second cluster with the high cases in all attributes, consisting of the USA, India, and Brazil.

Multi-criteria large group model based on consensus measure and subgroup leader mechanism for smart cities evaluation with hesitant information

Smart city refers to the use of various information technologies to integrate urban systems and services so as to improve the efficiency of resource utilisation and improve the quality of life for citizens. For many activities related to smart cities, such as the selection of pilot cities, a large number of experts from different functionalities or departments are usually invited to make evaluations of multiple attributes. The wide-spanning nature of smart cities needs cross-functional integration of various types of expertise. Therefore, it is necessary to develop a multi-criteria large group decision-making model to gather expert opinions from a wide range of sources to solve these problems. To do this, we first use the simple and fast algorithm for K-medoids clustering to classify experts into different subgroups and thereby reduce the complexity of the decision-making problem. Subgroup leaders will be selected at the same time to represent subgroups in subsequent decision-making processes. We then use the DEMATEL method to determine the weights of attributes. Next, to ensure that the decision outcome is supported by the majority of experts, a consensus-reaching process is proposed to reduce discrepancies in opinions. An illustrative example is adopted which involves the selection of pilot cities in Sichuan Province in order to verity the applicability of the model. Comparative analyses will be provided to verify the advantages of the proposed model. The results show that our model can effectively address evaluation problems associated with smart city activities involving a large group of experts.