K-medoids Algorithm Research Articles

Manhattan Distance-based K-Medoids Clustering Improvement for Diagnosing Diabetic Disease

Diabetes is a metabolic disorder characterized by blood glucose levels above normal limits. Diabetes occurs when the body is unable to produce sufficient insulin to regulate blood sugar levels. As a result, blood sugar management becomes impaired and there is no cure for diabetes. Early detection of diabetes provides an opportunity to delay or prevent its progression into acute stages. Clustering can help identify patterns and groups of diabetes symptoms by analyzing attributes that indicate these symptoms. In this study, researchers are using K-Medoid and Quantum K-Medoid methods for clustering diabetes data. Quantum computing utilizes quantum bits, or qubits, which can represent multiple states at the same time. Compared to classical computers, quantum computing has the potential for an exponential speedup in problem-solving. Researchers conducted a comparison between two methods: the classic K-Medoids method and the K-Medoids method utilizing quantum computing. The researchers found that both Quantum K-Medoid and Classic K-Medoid achieved the same clustering accuracy of 91%. In testing with the Quantum K-Medoids algorithm, it was found that the cost value in the 8th epoch showed a significant decrease compared to the Classical K-Medoids algorithm. This demonstrates that Quantum K-Medoid can be considered a viable alternative for clustering purposes.

Application Of Data Mining In Selecting Superior Products Using The K-Means And K-Medoids Algorithm Methods

As a supermarket, we are committed to always improving everything, including selecting the greatest goods. To evaluate which items are more superior or popular and which are less popular, you will want a sizable amount of information sources. To select products and identify those that belong in the superior product cluster, researchers employed the clustering method. The clustering strategy uses two forms of cluster analysis, k-means and k-medoids, which have related techniques. The research results show that the k-means algorithm's Davies Bouldin value is -0.430, whereas the k-medoids algorithm's Davies Bouldin value is -1.392. This suggests that the Davies Bouldin value of the k-medoids approach is the lowest, showing that the grouping findings of the k-means method are a better method to apply to the issue of choosing better products.

Unsupervised machine learning in financial anomaly detection: clustering algorithms vs. dedicated methods

The article presents the application of selected clustering algorithms for detecting anomalies in financial data compared to several dedicated algorithms for this problem. To apply clustering algorithms for anomaly detection, the Determine Abnormal Clusters Algorithm (DACA) was developed and implemented. This parameterized script (DACA) allows clusters containing anomalies to be automatically detected on the basis of defined distance measures. This kind of operation allows clustering algorithms to be quickly and efficiently adapted to anomaly detection. The prepared test environment has allowed for the comparison of selected clustering algorithms. K-Means, Hierarchical Cluster Analysis, K-Medoids, and anomaly detection: Stochastic Outlier Selection, Isolation Forest, Elliptic Envelope. The research has been carried out on real financial data, in particular on the income declared in the asset declarations of the targeted professional group. The experience of financial experts has been used to assess anomalies. Furthermore, the results have been evaluated according to a number of popular classification and clustering measures. The highest result for the investigated financial problem was provided by the K-Medoids algorithm in combination with the DACA script. It is worthwhile to conduct future research on the introduced solutions as an ensemble method.

Comparison of K-Means and K-Medoids Algorithms for Clustering Poverty Data in South Sumatra Using DBI Evaluation

Comparison of K-Means and K-Medoids Algorithms for Clustering Poverty Data in South Sumatra Using DBI Evaluation

The heating effect analysis of electromagnetic induction heating system based on the multi objective optimization strategy

Lithium-ion battery heating in cold weather is necessary to ensure its low-temperature performance and lifetime, so the multi objective optimization heating strategy based on the non-dominated sorting genetic algorithm II is introduced to improve the heating effect of electromagnetic induction heating system, in which the generated Pareto frontier as the research sample is input in K-medoid algorithm in order to reduce the feasible range of solution sets and then get the optimal induction coil parameters combination, and the temperature rise and terminal voltage output performance of battery are characterized by the electrochemical-thermal coupling model. Firstly, under the optimal parameters combination (N = 180, f = 9 kHZ, IA = 6 A), the battery can be heated from 243.15 K to 293.15 K in 1430 s with the temperature difference of 2.74 K, and the temperature field is mainly generated by the heat conduction from active material to mandrel during the heating process due to skin effect and material property. Secondly, the internal resistance is decreased by more than 3 times after the heating, giving rise to the substantial improvement of energy output ability. Finally, neither noticeable discharging capacity retention loss nor apparent voltage output difference is founded after the 200 repeated heating cycles, as compared with the normal charging-discharging cycle at 293.15 K, so the heating strategy has almost no effect on the battery aging and lifetime deterioration. Therefore, the proposed heating strategy can realize a suitable tradeoff among temperature-rising effect, lifespan and low-temperature operating performance, which has substantial potential to be a feasible method for improving the applicability of electric vehicles in a cold weather.

Edinburgh postpartum depression scores are associated with vaginal and gut microbiota in pregnancy

BackgroundPrenatal and postpartum depression may be influenced by the composition of host associated microbiomes. As such, the objective of this study was to elucidate the relationship between the human gut or vaginal microbiomes in pregnancy with prenatal or postpartum depression. Methods140 female participants were recruited at their first prenatal visit and completed the Edinburgh Postnatal Depression Scale (EPDS) to screen for depression and anxiety, in addition the EPDS was completed one month postpartum. Vaginal and stool biospecimens were collected in the third trimester, analyzed using 16S rRNA gene sequencing, and assessed for alpha and beta diversity. Individual taxa differences and clustering using the k-medoids algorithm enabled community state type classification. ResultsParticipants with higher postpartum EPDS scores had higher species richness and lower abundance of L. crispatus in the vaginal microbiota compared to those with lower EPDS scores. Participants with a higher prenatal EPDS score had lower species richness of the gut microbiome. Participants with a vaginal community state type dominated by L. iners had the highest mean prenatal EPDS scores, whereas postpartum EPDS scores were similar regardless of prenatal vaginal state type. LimitationsOur small sample size and participant's self-report bias limits generalizability of results. ConclusionsDepression in the prenatal and postpartum period is associated with the composition and diversity of the gut and vaginal microbiomes in the third trimester of pregnancy. These results provide a foundational understanding of the microbial relationships between maternal health and depression for identifying potential therapeutic treatments.

Time-series clustering of raw recordings of 24-hour ambulatory blood presssure for risk stratification in general population

Abstract Objective 24-hour ambulatory blood pressure measurements (ABPM) is the best out-of-office BP measurement used to refine cardiovascular (CV) risk stratification. Currently, only summary ABPM indexes are used in the clinical practice while the temporal characteristics of the raw BP recordings have not been explored. Therefore, in this study we employed an unsupervised machine learning (ML) algorithm suitable for multivariate time-series analysis to identify distinct BP and heart rate (HR) profiles associated with the incidence of adverse CV events. Design and Method In 1391 community-dwelling individuals (mean age, 46.9 years; 50.97% women), we acquired 24-hour ABPM and clinical data and collected CV events (n=399) on average 22 years later. We employed multivariate dynamic time warping and k-medoids algorithm on the raw recordings of systolic and diastolic BP and HR. To assess the clinical significance of the derived clusters, we compared the clinical characteristics and the incidence of adverse events. We validated the trained model using 24-hour ABPM obtained from external population cohort (n=1137). Results The silhouette score and Dunn index showed that the optimal number of clusters was 4. Across all clusters we observed significant difference in age and sex distribution. Cluster 4 had the worst CV risk factors profiling, with higher office BP and anti-hypertensive medication intake compared to the other clusters. We observed differences in the incidence of adverse events between clusters (Figure), with cluster 1 representing the lowest risk group and cluster 4 the highest. After adjusting for traditional CV risk factors, including office systolic BP, cluster 1 had the lowest risk (HR:0.92; 95%CI:0.84-0.99; P=0.036) and cluster 4 the highest (HR:1.13; 95%CI:1.03-1.24; P=0.006) as compared to the average population risk. The analysis of the external validation cohort revealed similar ABPM patterns and clinical characteristics. Conclusion Employing an unsupervised ML model on the raw ABPM recordings we identified clinically meaningful clusters associated with an increasing cumulative incidence of CV events. This ML analysis paves the way towards utilization of temporal BP and HR information and subsequently optimization of CV risk stratification.

Energy-efficient trajectory optimization algorithm based on K-medoids clustering and gradient-based optimizer for multi-UAV-assisted mobile edge computing systems

The mobile edge computing system supported by multiple unmanned aerial vehicles (UAVs) has gained significant interest over the last few decades due to its flexibility and ability to enhance communication coverage. In this system, the UAVs function as edge servers to offer computing services to Internet of Things devices (IoTDs), and if they are located distant from those devices, a significant amount of energy is consumed while data is transmitted. Therefore, optimizing UAVs’ trajectories is an indispensable process to minimize overall energy consumption in this system. This problem is difficult to solve because it requires multiple considerations, including the number and placement of stop points (SPs), their order, and the association between SPs and UAVs. A few studies in the literature have been presented to address all of these aspects; nevertheless, the majority of them suffer from slow convergence speed, stagnation in local optima, and expensive computational costs. Therefore, this study presents a new trajectory optimization algorithm, namely ITPA-GBOKM, based on a newly proposed transfer-based encoding mechanism, gradient-based optimizer, and K-Medoids Clustering algorithm to tackle this problem more accurately. The K-medoid clustering algorithm is used to achieve better association between UAVs and SPs since it is less sensitive to outliers than the K-means clustering algorithm; the transfer function-based encoding mechanism is used to efficiently define this problem’s solutions and manage the number of SPs; and GBO is utilized to search for the best SPs that could minimize overall energy consumption, including that consumed by UAVs and IoTDs. The proposed ITPA-GBOKM is evaluated using 13 instances with several IoTDs ranging from 60 to 700 to show its effectiveness in dealing with the trajectory optimization problem at small, medium, and large scales. Furthermore, it is compared to several rival optimizers using a variety of performance metrics, including average fitness, multiple comparison test, Wilcoxon rank sum test, standard deviation, Friedman mean rank, and convergence speed, to show its superiority. The experimental results indicates that this algorithm is capable of producing significantly different and superior results compared to the rival algorithms.

Comparative Analysis of K-Means and K-Medoids Algorithms for Product Sales Clustering and Customer

In today's rapidly evolving business landscape, effective product management is crucial for maintaining a company's competitive advantage. Comprehensive analysis is essential for providing insights that inform strategic business development decisions. This study examines the sales data of PT XYZ from July 2020 to May 2024 using the K-Medoids algorithm, with dimensionality reduction applied through Principal Component Analysis (PCA). The clustering results identified three customer segments: Cluster 1 with 46 customers, Cluster 2 with 76 customers, and Cluster 3 with 62 customers. For product segmentation, four clusters were identified: Cluster 1 with 52 products, Cluster 2 with 12 products, Cluster 3 with 20 products, and Cluster 4 with 53 products. The K-Medoids algorithm demonstrated superior performance compared to K-Means in terms of cluster separation and interpretability, with visualizations that enhance the understanding of customer and product distributions. This research aids the company in enhancing customer satisfaction, optimizing inventory, and increasing profitability.

Quantum-Enhanced K-Medoids Clustering: Comparative Analysis of Stroke Medical Data

Stroke is a severe medical condition that occurs when the blood supply to parts of the brain is interrupted or reduced, resulting in brain tissue that lacks oxygen and nutrients. This causes brain cells to start to die in minutes. Early prevention reduces the risk of stroke. In this study, a quantum computing approach is used to improve the performance of the K-Medoids method. A comparative analysis of these methods was carried out with a focus on their performance, especially on the accuracy of the test results. The investigation was carried out using a data set of stroke patient medical records. The data set was tested using the classical and K-Medoids methods with a quantum computing approach utilizing Manhattan distance calculations. The findings of this research reveal improvements in the K-Medoids algorithm with Manhattan distance calculation influenced by the integration of a quantum computing framework. In particular, the simulation test results show an increase in accuracy from the classical K-Medoids method to the K-Medoids method with a quantum computing approach, from 52% to 64%. These results highlight that the performance of the K-Medoids method with a quantum computing approach is superior to that of the classical K-Medoids method.

Analyzing and Optimizing the Distribution of Blood Lead Level Testing for Children in New York City: A Data-Driven Approach.

This study investigates blood lead level (BLL) rates and testing among children under 6 years of age across the 42 neighborhoods in New York City from 2005 to 2021. Despite a citywide general decline in BLL rates, disparities at the neighborhood level persist and are not addressed in the official reports, highlighting the need for this comprehensive analysis. In this paper, we analyze the current BLL testing distribution and cluster the neighborhoods using a k-medoids clustering algorithm. We propose an optimized approach that improves resource allocation efficiency by accounting for case incidences and neighborhood risk profiles using a grid search algorithm. Our findings demonstrate statistically significant improvements in case detection and enhanced fairness by focusing on under-served and high-risk groups. Additionally, we propose actionable recommendations to raise awareness among parents, including outreach at local daycare centers and kindergartens, among other venues.

Identification and Spatial Characterization of suburban areas in Chengdu

As transitional territories where urban and rural functions interpenetrate, suburban areas have multiple values such as recreation, ecology, agriculture and landscape. Promoting their benign development is the key to realizing global sustainable development of urban and rural areas. Planning control is based on precisely defining the spatial extent of suburban areas, and understanding suburban area differentiation and its driving forces scientifically is a key component in raising the bar for focused planning. However, there are currently few research on fine-scale quantitative identification of suburban regions because of their spatial complexity, ambiguous boundaries, and structural dynamic features. Based on this, this paper develops a multi-dimensional identification index system for suburban areas by using multi-source big data and remote sensing information technology, employs the random forest model and the K-Medoids clustering algorithm, identifies the distribution of suburban areas and their subdivided types, analyzes spatial differentiation characteristics, and conducts empirical research using Chengdu City as an example. The findings demonstrate that: (1) The majority of suburban areas in Chengdu City are concentrated in groups or belts surrounding the urban centers of various districts and counties, and the higher the functional class of a district or county, the denser the distribution of suburban areas. (2)The distribution pattern of suburban areas in Chengdu basically conforms to its planned urban hierarchy system.(3)Urban-oriented, suburban-interacted and rural-oriented suburban areas have their own characteristics in land use level, economic development and regional population characteristics.(4)Regional openness, government behavior, social development and geographical location are the core driving factors of spatial differentiation of suburban areas in Chengdu, and the interaction between urban and rural systems, the flow of social resources and the transformation of development models are the spatial mapping dimensions that affect its differentiation.(5)Chengdu can draw up detailed regulatory planning of suburban units to standardize and guide the overall development of suburban areas. In support of the global sustainable development goals, this study offers a methodology for precisely and impartially defining suburban areas, assisting in the implementation of integrated urban-rural development globally.

Building electricity load forecasting based on spatiotemporal correlation and electricity consumption behavior information

Accurate prediction of building electricity load is essential for grid management and building optimization operations. This paper proposes a novel approach based on spatiotemporal correlations and electricity consumption behavior information. The K-Medoids algorithm and the Derivative Dynamic Time Warping (DDTW) distance are employed to explore the correlation between electricity consumption behaviors among different partitions and floors. Different partitions and floors are clustered and grouped, followed by modifying the adjacency matrix with electricity consumption behaviors. The hybrid model and K-Medoids-LSTM model are proposed separately for clusterable nodes and non-clustered nodes. For clusterable nodes, spatial-temporal features are extracted, trained, and predicted with the hybrid model based on graph neural networks (GNNs) and LSTM models. A K-Medoids-LSTM model based on the K-Medoids algorithm is proposed to predict the electricity load of the non-clustered nodes. To explore the model's practicality, we predicted the building electrical load under different dataset sizes. The model achieves an R2 above 0.89, and the MAE, MSE, and RMSE of the GCN-LSTM and GAT-LSTM models all remain below 0.1, indicating strong predictive capabilities. The results demonstrate that, without relying on other external features, the proposed method can accurately predict the building electricity load for different partitions and floors simultaneously.

Pengelompokkan Kualitas Pemuda Menurut Provinsi di Indonesia Tahun 2023 Menggunakan Partition Clustering

The percentage of youth in Indonesia in the last decade has been in the range of 23-24 percent of the Indonesian population. Youth have an active role in national development. Therefore, youth development is the focus of the government because it is the key to accelerating development. This study aims to categorize provinces based on the condition of youth in health, education, and socio-economic aspects in 2023 so that it can assist the government in determining youth development policies effectively and inclusively. Based on the validation, k-medoids algorithm was chosen as the best clustering method with the number of clusters as many as 4. In general, the four clusters have quite different characteristics. Youth development in education, socio-economic, and health aspects by province still shows gaps. Therefore, equalization efforts tailored to the needs of provincial clusters are needed to achieve more equitable and inclusive progress.

Stock Grouping Based on Price Earnings Ratio and Price Book Value Using K-Medoids Algorithm

Investing involves allocating funds to achieve optimal returns by evaluating opportunities and managing risks in asset acquisition. Recently, many news reports have highlighted issues in the Indonesian capital market, such as stock investors using online loan funds for trading, which often leads to debt. This research aims to apply the K-Medoids algorithm for stock clustering, enabling investors to select fundamentally sound stocks based on the Price-Earnings Ratio (PER) and Price-Book Value (PBV). The K-Medoids method results show that Cluster 1 includes 93 stocks with moderate PER and PBV values. Cluster 2 comprises 91 stocks with the lowest PER and PBV values. Cluster 3 contains 113 stocks with the highest PER and PBV values. Developing an information system that classifies stocks based on PER and PBV can help investors analyze and make investment decisions more effectively.

K-Medoids and Support Vector Machine in Predicting the Level og Building Damage in Earthquake Insurance Modeling

Yogyakarta, an Indonesian province prone to earthquakes, frequently suffers extensive damage to buildings, necessitating insurance coverage to mitigate potential losses. This study aims to forecast earthquake insurance premiums by predicting building damage levels resulting from earthquakes. Utilizing data from buildings affected by the June 30, 2023, earthquake in Yogyakarta, we employ K-Medoids Clustering and Support Vector Machine (SVM) to predict two categories of building damage: minor (labelled as 1) and heavy (labelled as 2). The total premiums for minor damage range from approximately USD 86.55 to USD 288.50, while for heavy damage, they range from USD 120.05 to USD 400.18 using the K-Medoids algorithm. Meanwhile, premiums for minor damage range from USD 83.14 to USD 277.13, and for heavy damage, they range from USD 223.67 to USD 745.55 using the SVM algorithm.

Tumor Heterogeneity in Gastrointestinal Cancer Based on Multimodal Data Analysis.

Gastrointestinal cancer cells display both morphology and physiology diversity, thus posing a significant challenge for precise representation by a single data model. We conducted an in-depth study of gastrointestinal cancer heterogeneity by integrating and analyzing data from multiple modalities. We used a modified Canny algorithm to identify edges from tumor images, capturing intricate nonlinear interactions between pixels. These edge features were then combined with differentially expressed mRNA, miRNA, and immune cell data. Before data integration, we used the K-medoids algorithm to pre-cluster individual data types. The results of pre-clustering were used to construct the kernel matrix. Finally, we applied spectral clustering to the fusion matrix to identify different tumor subtypes. Furthermore, we identified hub genes linked to these subtypes and their biological roles through the application of Weighted Gene Co-expression Network Analysis (WGCNA) and Gene Ontology (GO) enrichment analysis. Our investigation categorized patients into three distinct tumor subtypes and pinpointed hub genes associated with each. Genes MAGI2-AS3, MALAT1, and SPARC were identified as having a differential impact on the metastatic and invasive capabilities of cancer cells. By harnessing multimodal features, our study enhances the understanding of gastrointestinal tumor heterogeneity and identifies biomarkers for personalized medicine and targeted treatments.

Research on Outlier Detection Methods for Dam Monitoring Data Based on Post-Data Classification

Safety monitoring of hydraulic structures is a critical task in the field of hydraulic engineering construction. This study developed a method for preprocessing and classifying monitoring data for the identification of gross errors in hydraulic structures. By utilizing linear regression and wavelet analysis techniques, it effectively differentiated various waveform characteristics in data sets, such as Sinusoidal Wave Cyclical, Triangular Wave Cyclical, Seasonal Cyclical, and Weakly Cyclical growth types. In the experiments for gross error identification, the 3σ algorithm, K-medoids algorithm, and Isolation Forest algorithm were applied to test the data. The results showed that the K-medoids algorithm excelled in processing Sinusoidal Wave Cyclical Data Sets; the 3σ algorithm adapted better to Triangular Wave Cyclical Data Sets; the Isolation Forest algorithm performed well in handling data sets with significant anomalies or atypical fluctuations and excelled in scenarios with strong seasonality and large data fluctuations; and for complex Weakly Cyclical Growth Data Sets, all three algorithms were less effective, indicating the potential need for more advanced analysis methods or a combination of multiple techniques. Testing on actual engineering data further confirmed the importance of using specific gross error identification techniques for special data types after data set pre-classification, providing a more effective technical solution for the safety monitoring of hydraulic structures.

K-medoid clustering containerized allocation algorithm for cloud computing environment

Load balancing is critical for container-based cloud computing environments for several reasons. A lack of appropriate load balancing techniques could result in a decrease in performance and possible service interruptions as some nodes get overloaded, while others are left underutilized. Cloud service providers can reduce latency and boost system performance by strategically placing containers using clustering algorithms. These techniques aid in efficiently using resources and load balancing by clustering related containers together according to their shared attributes. Clustering strategies are effective in allocating and controlling resources to meet the demands of a changing workload. Algorithms for clustering combine related workloads or containers into clusters, improving performance isolation and maximizing resource usage. One popular methodology for data clustering is the K-Medoid Clustering Algorithm. It is especially helpful when working with categorical data or when the dataset contains outliers. K-medoids is an unsupervised clustering approach where the core of the cluster is made up of data points known as “medoids.” A medoid is a location in the cluster whose total distance to every object in the cluster—also known as its dissimilarity—is as small as possible. Any appropriate distance function may be used, such as the Manhattan distance, the Euclidean distance, or another one. Thus, by choosing K medoids from our data sample, the K-medoids method splits the data into K clusters. This work presents the K-Medoid clustering technique for containers, which can enhance load balancing, decrease resource execution times, and increase resource utilization rates all at the same time. The results of the experiment show that the proposed method outperforms MACO and FCFS in terms of throughput by about 70% when number of cloudlets increased. The relative improvement of execution time of the proposed K-medoid algorithm w.r.t FCFS is about 50%.

SCG variability and spectral energy distribution during normal breathing and breath hold at different lung volumes and airway pressures

Seismocardiographic (SCG) signals are chest wall vibrations induced by cardiac activity and are potentially useful for cardiac monitoring and diagnosis. SCG waveform is observed to vary with respiration, but the mechanism of these changes is poorly understood as alterations in autonomic tone, lung volume, heart location and intrathoracic pressure are all varying during the respiratory cycle. Understanding SCG variability and its sources may help reduce variability and increase SCG clinical utility. This study investigated SCG variability during breath holding (BH) at two different lung volumes (i.e., end inspiration and end expiration) and five airway pressures (i.e., 0, ± 2–4, and ± 15–20 cm H2O). Variability during normal breathing was also studied with and without grouping SCG beats into two clusters of similar waveform morphologies (performed using the K-medoid algorithm in an unsupervised machine learning fashion). The study included 15 healthy subjects (11 Females and 4 males, Age: 21 ± 2 y) where SCG, ECG, and spirometry were simultaneously acquired. SCG waveform variability was calculated at each experimental state (i.e., lung volume and airway pressure). Results showed that breath holding was more effective in reducing the intra-state variability of SCG than clustering normal breathing data. For the BH states, the intra-state variability increased as the airway pressure deviated from zero. The subaudible-to-audible energy ratio of the BH states increased as the airway pressure decreased below zero which may be related to the effect of the intrathoracic pressure on cardiac afterload and blood ejection. When combining the BH waveforms at end inspiration and end expiration states (at the same airway pressures) into one group, the intra-state variability increased, which suggests that the lung volume and associated change in heart location were a significant source of variability. The linear trend between airway pressure and waveform changes was found to be statistically significant for BH at end expiration. To confirm these findings, more studies are needed with a larger number of airway pressure levels and larger number of subjects.