Cluster Analysis Model Research Articles

Segmented assimilation trajectories of physician trust among internal migrants in Shanghai, China: A cross-sectional study

Physician trust is necessary for improving physician–patient relationships and maintaining an effective health-care system. Most studies on the theory of segmentation assimilation have mainly focused on comparing immigrants in different countries and few studies have applied segmented assimilation to clarify physician–patient relationships, especially physician trust among internal migrants in the context of Chinese megacities. A random sample of 1200 internal migrants was collected through an online questionnaire conducted in Shanghai from August to December 2021. An exploratory K-means cluster analysis and multivariate logistic regression models were employed to identify patterns of segmented assimilation and examine their relationships with physician trust, as well as the factors influencing physician trust among internal migrants in Shanghai. Results show four main patterns were revealed, namely first-generation classic assimilation, first-generation integrative assimilation, first-generation segmentation, and second-generation underclass assimilation, supporting the theory of segmented assimilation. Association between assimilation pattern and physician trust was observed. A relationship was found between assimilation patterns and trust in physicians. Migrants belonging to the first-generation classic and integrative assimilation groups exhibited a higher level of trust in doctors compared to those in the segmented assimilation groups. Additionally, undergraduate and postgraduate education attainment, an annual income of <400 000 yuan, visiting the physician ≥2 times in the past year, and self-rated health status before and after population floating were significant contributing factors for good physician trust. The government can implement corresponding measures to maintain physician trust, improve cultural adaptability instead of only at the economic level among internal migrants, achieving the facilitation of their understanding of the health system and health service utilization to increase physician trust.

Experimental study on acoustic emission damage in precast reinforced concrete interior joints containing disc springs

To address the inferior performance of precast reinforced concrete beam-column joints compared to cast-in-place joints during earthquakes, this study designed a precast concrete frame beam-column interior joint incorporating disc springs. The disc springs, installed based on grouted sleeve connections, aimed to enhance structural ductility and mitigate seismic forces on the joints. Five joints, including one cast-in-place joint, two ordinary precast joints, and two precast joints with disc springs, were tested quasi-statically and monitored using acoustic emission technology. The precast joints had axial compression ratios of 0.2 and 0.4. Parameters such as energy, count, and b-value from acoustic emission monitoring were analyzed. The Gaussian mixture model was employed for the clustering analysis of RA-AF parameters, and a damage model was developed to evaluate performance differences among the joints in the quasi-static tests. The results indicate that precast joints with the same axial compression ratio can achieve performance similar to cast-in-place joints, with more rapid damage development at higher axial compression ratios. Installing disc springs decelerates damage progression in the later stages of loading, maintaining joint integrity for a certain period even after reaching maximum load, thereby demonstrating improved ductility. At higher axial compression ratios, the b-value curve of joints with disc springs more closely resembles that of ordinary joints. Disc springs' positive effect is more pronounced in clustering analysis and damage models, such as their lower proportion of shear fractures.

Mental health issues of higher education students reflected in academic research: A text mining study

Objective: This study investigated mental health issues among higher education students to identify key concepts, topics, and trends over three periods of time: Period 1 (2000–2009), Period 2 (2010–2019), and Period 3 (2020–May 2024). Methods: The study collected 11,732 bibliographic records from Scopus and Web of Science, published between January 2000 and May 2024, and employed textual analysis methods, including keyword co-occurrence analysis, cluster analysis, and topic modeling. Results: In Period 1, general topics related to mental health disorders and treatments were identified. Period 2 showed prominence of well-being and help-seeking, as well as the emergence of digital mental health. Period 3 emphasized the impact of COVID-19 and increased technology usage. Conclusions: Based on the findings, we discussed the significance of the study and practical implications for clinicians and policymakers, as well as methodological implications for researchers. Additionally, the limitations of the study and future research were addressed.

Pilot Study on the Use of Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy for Diagnosing and Characterizing Cardiac Amyloidosis.

Amyloidosis diagnosis relies on Congo red staining with immunohistochemistry and immunofluorescence for subtyping but lacks sensitivity and specificity. Laser-microdissection mass spectroscopy offers better accuracy but is complex and requires extensive sample preparation. Attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy offers a promising alternative for amyloidosis characterization. Cardiac tissue sections from nine patients with amyloidosis and 20 heart transplant recipients were analyzed using ATR-FTIR spectroscopy. Partial least squares discriminant analysis (PLS-DA), principal component analysis (PCA), and hierarchical cluster analysis (HCA) models were used to differentiate healthy post-transplant cardiac tissue from amyloidosis samples and identify amyloidosis subtypes [κ light chain (n = 1), λ light chain (n = 3), and transthyretin (n = 5)]. Leave-one-out cross-validation (LOOCV) was employed to assess the performance of the PLS-DA model. Significant spectral differences were found in the 1700-1500 cm-1 and 1300-1200 cm-1 regions, primarily related to proteins. The PLS-DA model explained 85.8% of the variance, showing clear clustering between groups. PCA in the 1712-1711 cm-1, 1666-1646 cm-1, and 1385-1383 cm-1 regions also identified two clear clusters. The PCA and the HCA model in the 1646-1642 cm-1 region distinguished κ light chain, λ light chain, and transthyretin cases. This pilot study suggests ATR-FTIR spectroscopy as a novel, non-destructive, rapid, and inexpensive tool for diagnosing and subtyping amyloidosis. This study was limited by a small dataset and variability in measurements across different instruments and laboratories. The PLS-DA model's performance may suffer from overfitting and class imbalance. Larger, more diverse datasets are needed for validation.

Segmenting the potential users of MaaS by combining latent class cluster analysis and structural equation modeling

To reduce problems caused by the excessive use of private vehicles, transportation planners have been actively looking for sustainable transport related solutions. Mobility as a Service (MaaS) is gaining popularity due to its capability to integrate transport opportunities into one platform. Studies on travelers’ acceptance of MaaS have been conducted, but there remains a gap in understanding heterogeneity regarding intentions of using MaaS particularly concerning attitudinal factors. Based upon the extension of the unified theory of acceptance and use of technology, we plan to reveal travelers’ segmentation by using latent class cluster analysis and to examine path heterogeneities in the identified traveler clusters by applying a structural equation modeling. The results confirm that the traveler clusters consider the importance of the relatives and media's viewpoints toward MaaS, while the car user cluster is more interested in adopting MaaS if the service could promise enough flexibility of traveling in terms of time and destination. The results related to the heterogeneity in transport mode choice are useful for constructing MaaS packages. This study could be advantageous for MaaS operators as it enables them to understand some attitudinal factors influencing willingness to take up MaaS thus realizing suitable, effective promotions.

Clusters of Sociodemographic Characteristics and Their Association with Food Insecurity in Mexican University Students.

Food insecurity (FI) expresses irregular access to sufficient, safe, and nutritious food. We analyze the association between clusters of sociodemographic characteristics and FI in university students from Mexico. The Latin American and Caribbean Food Security Scale was applied, and household type, socioeconomic status (SES), and the characteristics of the head of the household, among others, were asked in a cross-sectional study. We performed Two-Step cluster analysis and logistic regression models. We included 466 university students (72.5% women, 47% with FI). We identified three clusters; Cluster-1 (n = 163) included a single-parent (37.4%) or co-resident/roommate (27%) household type, middle SES (63.2%); the head of the household is usually a woman (76.1%), single (71.8%), and with bachelor's degree (41.7%) or less educational level (46.6%). Cluster-2 (n = 144) included the nuclear (89.6%) household type, high SES (76.4%); the head of the household is usually a man (79.2%), in a relationship (99.3%), and with a bachelor's degree (63.2%) or graduate level (33.3%). Cluster-3 (n = 147) is very similar to Cluster-2. The difference is that Cluster-3 includes middle SES (70.1%), and the head of the household's educational level is high school or less (100%). Only belonging to Cluster-1 was positively associated with FI (OR = 1.96, 95%CI: 1.24, 3.09). These results show the interrelationships between multiple sociodemographic characteristics and should contribute to developing interventions that simultaneously address numerous sociodemographic factors to diminish FI in university students.

Clusters of Relative Deprivation for Children under 12 Years of Age

ABSTRACT We compare children according to whether and to what extent they are deprived using the data from the German family survey “Growing up in Germany (in short: AID:A)” 2019. We classify the children into different types of relative deprivation using a Ward cluster analysis and use multinomial logistic regression models to estimate each child’s probability of being assigned to each type. Our results indicate the prevalence of relative deprivation of children covaries with economic and non-economic resources available to the children’s households. Further, we find only marginal differences between children within the same household.

Application of Business Information Management in Cross-border Real Estate Project Management

Cross-border real estate project management is inherently challenging due to its complexity and diversity. This study investigates the efficacy of business information management systems (BIMS) in managing such projects and employs machine learning models for performance prediction analysis. Utilizing data from 250 valid questionnaires and 15 in-depth interviews, multiple regression analysis, classification algorithms, and clustering analysis models were applied. The results indicate that system quality, information quality, and service quality significantly enhance project management efficiency and user satisfaction. Specifically, the adoption of BIMS reduces average project completion time and cost overrun rates, thereby improving management effectiveness. Commercial real estate projects reported the highest average investment at $70 million, mixed-use projects exhibited the longest average completion time of 25 months, and residential real estate projects achieved the highest management efficiency score, averaging 8.0. The regression model's coefficient of determination (R²) was 0.68, the classification model achieved an 85% accuracy in identifying risk factors, and clustering analysis categorized projects into high-efficiency management, risk-concentrated, and resource-intensive types. These findings underscore the substantial value of BIMS in cross-border real estate project management, providing robust management tools and decision support. However, the study's limitations include a small sample size and restricted data sources. Future research should aim to expand the sample size and incorporate more diverse data sources to enhance the findings' generalizability and accuracy.

Multiple spatial-temporal scales assessment of solar and wind resources potential integrating geospatial-technology-correlation indicators: A case study of Hunan Province

Achieving the green transformation of energy structure requires extensive utilization of renewable energy, especially wind and solar energy. However, spatial distribution and correlation between solar and wind resources in provincial region are unclear, hinders their further development. To accurately assess provincial solar and wind potential, a multiple spatial and temporal assessment model based on geographic information system is proposed, which integrates geographic constraint analysis, resource potential quantification, correlation analysis, cluster analysis and gravity model. The model forms a more comprehensive evaluation of renewable energy, including resource capacity and distribution at multiple temporal and spatial scales, resource-rich area distribution, multi-energy correlation characteristic and regional energy interaction potential. The results showed that geographical potentials for photovoltaic and wind power in Hunan Province are 27698.73 km2 and 35013.38 km2, with an annual average of 608903.23 GWh and 187544.84 GWh, respectively. The highest contributions of resource potential are Yongzhou, Hengyang and Yueyang. Hengyang exhibits strong potential for supplying neighboring cities. Solar and wind energy in Hunan display a certain spatial synergy and complementarity, but weak correlation in hourly and monthly scales. The evaluation model can provide theoretical and practical guidance for renewable energy macro-analysis, project planning and precise decision-making.

Research on Classification and Identification of Glass Relics Based on Pearson Clustering and Machine Learning Models

Ancient glass is very susceptible to weathering due to the influence of buried environments. In order to classify and identify glass artifacts, this article uses the Pearson cluster analysis model to classify them into four types of high potassium and three types of lead and barium. The chemical cost composition of the classification types is determined, and then the composition of each chemical component is accumulated and arranged. The proportion of each type's content is used as the basis for subclassing. Finally, multiple machine learning methods were used to identify unknown types of glass artifacts, and the results showed that the accuracy of the support vector machine model reached 100%, with a relative improvement of 2% -11%. Therefore, the support vector machine model was considered for classification.

Automatic pricing and replenishment decision-making for vegetable products based on optimization models

Addressing the replenishment and pricing issues of vegetable products is crucial for ensuring product quality and freshness, optimizing sales combinations, refining pricing strategies, and enhancing operational efficiency. Through scientific data analysis and decision-making, supermarkets can better meet consumer demands, enhance competitiveness, and achieve sustainable development. This paper discusses the complex issues of procurement and pricing of fresh vegetable products in current supermarkets. It employs methods such as hierarchical clustering analysis, Topsis evaluation, and optimization models to construct data models, establishing multiple models to address replenishment and pricing decision-making from various perspectives. The research indicates, firstly, the paper categorizes vegetable products into four clusters, explores complementary and substitute products within them, and discovers that reasonable sales combinations among different types of single products can mutually promote sales, leading to higher economic benefits for supermarkets. Secondly, the paper derives a mathematical model describing the relationship between total profit, total sales volume of individual products, and pricing. This model provides valuable recommendations for supermarkets replenishment and pricing decisions, ensuring practical implementation of pricing and replenishment plans. Thirdly, the paper establishes a model for maximizing profits under constant replenishment quantities, assisting supermarkets in formulating more scientific replenishment plans for individual products within a limited number of available items. By judiciously applying the innovative mathematical models presented in this paper, supermarkets can obtain reliable market analysis and make corresponding replenishment and pricing decisions.

Nurses' Personality Traits and Perceived Work Environments During Public Health Emergencies: Implications for Nursing Workforce Planning.

The nursing workforce faces substantial challenges, particularly in the aftermath of the COVID-19 era. Developing an effective strategy for workforce maintenance and the strategic deployment of nurses is crucial. This study aimed to explore and categorize nurses' personality traits, with a focus on analyzing differences in their perceptions of the nursing work environment. Between January 2023 and February 2023, a multi-center cross-sectional study was carried out involving nurses from 12 tertiary hospitals actively engaged in frontline COVID-19 response duties. Through cluster sampling, surveys were distributed among eligible nursing staff, comprising a general information questionnaire, the Chinese Big Five Personality Questionnaire-Short Form, subjective evaluations of emergency nursing management, and the Chinese Nursing Work Environment Scale for Public Health Emergencies. Various statistical analyses, such as descriptive analysis, cluster analysis, non-parametric tests, and general linear model analysis, were employed to investigate the correlation between personality types and the perception of nursing work environments. The analysis encompassed 1059 valid questionnaires, reflecting the experiences of frontline nurses. The majority of these nurses possessed 1-5 years of experience, held junior professional titles, volunteered for their roles, and served as attending nurses. Categorization based on personality traits revealed three groups: resilient (35.60%), ordinary (16.15%), and distressed (48.25%) types. Significantly distinct perceptions of nursing work environments emerged among these categories, with resilient and ordinary types expressing notably higher satisfaction compared to the distressed group (H value = 256.487, p < 0.001). This study illustrates the connection between nurses' perceived working environment and their personality traits. Nursing managers should factor in nurses' personality traits when choosing and deploying frontline responders during public health emergencies. Prioritizing resilient-type nurses and crafting a supportive work environment that aligns with nurses' characteristics is indispensable for an effective emergency response.

Catalytic ozonation of 2, 4-dichlorophenoxyacetic acid wastewater by Fe-La@ZE catalyst

This study developed a salt-tolerant multimetal ozone catalyst Fe-La@Zeolite (Fe-La@ZE) and effectively catalyzed the ozone depletion of high-salinity 2,4-dichlorophenoxyacetic acid wastewater (2,4-D) and an entropy to evaluate the comprehensive gray cluster analysis model was created using catalytic ozonation conditions. Characterization results show that Fe-La@ZE has a developed pore structure and Fe and La are successfully loaded into the interior and surface of the zeolite support, which are in the form of α-Fe2O3 and La2O3. The COD removal rate of zeolite adsorption, ozone oxidation alone and catalyzed ozone oxidation with empty zeolite support are 4.9 %, 46.5 % and 57.5 %. The removal rates of COD, total phenol and 2,4-dichlorophenol by catalytic ozonation under the conditions of pH = 0.89, ozone dose of 0.2 g/L, catalyst filling rate of 10 % and height to diameter ratio of the reactor 6:1 are 84.2 %, 87.8 % and 99.0 % respectively. Hydroxy radicals (OH) and superoxide radicals (O2) are the main active substances that catalyze the degradation of 2,4-dichlorophenol in the ozone oxidation system. Water quality characterization shows that most phenolic organic compounds and macromolecular organic acids are broken down and converted into intermediate products. The comprehensive analysis model of entropy gray clustering is used to evaluate the operating condition data of the catalytic ozonation treatment of 2,4-dichlorophenoxyacetic acid wastewater with the Fe-La@ZE catalyst and determine the optimal operating condition.

Chilean Market Protein Shakes Composition.

Understanding the nutritional content of protein supplements is crucial for optimal nutritional planning among athletes and other people. Distribution of macronutrients and aminograms in the main products available in the national Chilean market remains unknown. A descriptive cross-sectional study was conducted to identify the main protein supplements available in the Chilean market. Information on macronutrients and aminograms from the nutritional labels of each product was extracted. The analysis considered the content per portion and per 100 g. Cluster analysis models and graphical representations were explored. Eighty protein shakes were assessed in the Santiago de Chile market. The median protein dosage was 32 g (range from 25 to 52), and the median energy value stood at 390 kcal (range from 312 to 514). The median protein content per 100 g of product was found to be 75 g (range from 42.5 to 97.2). The combined median concentration of amino acids was 4749.75 mg. Among these, the essential amino acid L-Tryptophan exhibited the lowest concentration at 1591.50 mg, while the conditional amino acid L-Glutamine had the highest median concentration at 17,336 mg. There was a significant prevalence of animal-derived products, placing specific emphasis on protein supplements that feature elevated levels of the amino acids L-Glutamine and L-Leucine.

Sugar-Free Herbal Beverage Preferences: Customer Clustering Insights

Amidst the increased demand for public health products caused by the COVID-19 epidemic, Indonesia's herbal business has recognized opportunity for expansion, particularly in the area of sugar-free instant herbal beverages. The purpose of this study is to define customer profiles and identify the factors impacting customers' willingness to pay for such products using a dual technique of cluster analysis and Structural Equation Modeling - Partial Least Squares (SEM-PLS). The study identifies three unique client groups: teen students (Cluster 1), early adult workers (Cluster 2), and late adult workers (Cluster 3). Income, health consciousness, product quality, and price are factors that have a favorable and significant impact on willingness to pay across all clusters. Notably, in Cluster 3, willingness to pay is more influenced by product quality, health consciousness, and price. These findings are useful for the Indonesian herbal business, as they will guide targeted marketing tactics and product development efforts to efficiently serve to varied consumer demographics.

Moral foundations in gender violence cases decided in Portuguese courts

Gender violence encompasses a multitude of morally problematic psychological, physical, and sexual behaviors that, in most countries, constitute criminal offenses. In this study, we investigate the association between moral foundations ( Care, Fairness, Loyalty, Authority, and Sanctity) and punitive responses to gender violence offenses. Our case study focuses on gender violence in Portugal, a country in which these offenses are a prevalent social problem. We collected data on gender violence legal cases decided in Portuguese courts between 2002 and 2022, and we used a latent class cluster analysis model to identify the complex patterns in the data and reduce such patterns to a distinct number of clusters. Four main clusters unravel latent relations between the foundations mapped in the legal narratives and corresponding punitive responses: (i) Affirmative with suspended prison time (moral rhetoric rooted in Authority); (ii) Mixed outcomes but no prison time (moral rhetoric rooted in Sanctity); (iii) Affirmative with lengthy prison time large compensation (moral rhetoric rooted in Loyalty and Care); and (iv) Affirmative with court fines (moral rhetoric rooted in Fairness). The moral foundations provide a valuable lens to understand the problem of gender violence, but further research is needed to establish the causal mechanisms between morality and punitive responses to gender violence.

Prognostic correlation between specialized capillary endothelial cells and lung adenocarcinoma

BackgroundIn-depth analysis of the functional changes occurring in endothelial cells (ECs) involved in capillary formation can help to elucidate the mechanism of tumour vascular growth. MethodsAppropriate datasets were retrieved from the GEO database to obtain single-cell data on LUAD samples and adjacent normal tissue samples. ECs were selected by an automatic annotation program in R and further subdivided based on reported EC marker genes. Functional changes in different types of capillary ECs were then visualized, and the concrete expression was classified by genetic data in the TCGA. Finally, a prognostic model was constructed to predict immunoinfiltration status, survival and drug therapy effects. ResultsThe LUAD data contained in the GSE183219 dataset were suitable for our analysis. After dimensionality reduction analysis and cell annotation, EC general capillary and EC aerocyte subsets as capillary specialized phenotypes showed a series of functional changes in tumour samples, with a total of 108 genes found to undergo functional changes. Use of CellPhoneDB revealed a close interaction of activity between ECs. After integration of TCGA, GSE68465 and GSE11969 datasets, the genes obtained were analysed by cluster analysis and risk model construction, identifying 8 genes. Drug sensitivity, immune cell and molecular differences can be accurately predicted. ConclusionsEC general capillary and EC aerocyte subsets are recognized capillary ECs in the tumour microenvironment, and the functional changes between them are relevant to the prognosis and treatment of LUAD patients and have the potential to be used in target therapy.

Dielectric spectroscopy with machine learning integrated approach to analysis the influence of N:P ratio on algae growth modeling for forecasting

The task of monitoring algal biomass is a significant challenge in contemporary environmental conservation. A novel technique called dielectric spectroscopy with machine learning integration is an intelligent approach to address this challenge. This work optimized a machine learning supervised PCA, Hierarchical cluster analysis & linear regression model -strengthen the real-time dielectric spectroscopy approach to study the influence of N:P ratios on a model algae Desmodesmus sp. growth for forecasting. As a primary feature, different N:P concentrations (1:1, 2:1, 3:1, 1:2, and 1:3) were utilized to characterize the relationship between dielectric (S11(dB) & ε'') and conventional UV–vis spectroscopy properties (OD686 & Chl-a). The dielectric measurements of S11(dB) & ε'' are consistent with the OD686 & Chl-a value. Highest Desmodesmus sp. OD686 0.87, 0.76 recorded at N:P 3:1, 2:1 ratio respectively. The machine learning based PCA, Hierarchical cluster analysis and linear regression model were applied and found at 0–8 days incubation Desmodesmus sp. growth biomass negatively corelated with S11(dB) & ε“ and positively correlatedly with OD686 & Chl-a respectively, with R2 values > 0.95. While for the 0&8days measured ε”& S11(dB) were −6.169&-0.4765; −7.956 &-0.6775 respectively, the linear regression model was applied to predicted OD686 for 0&8 days was 0.299, 0.234; 0.817, 0.716 respectively, at the ratios N:P (3:1) and the RMES were 0.0047 & 0.0319 respectively. This analysis revealed that the machine learning based integrated approach has high accuracy. Thus, Dielectric spectroscopy and machine learning-integrated method could strengthen to the real-time monitoring of algal biomass or predicting to control the N:P ratios to protecting aquatic ecosystem.

Fatal Crash Occurrence Prediction and Pattern Evaluation by Applying Machine Learning Techniques

Background Highway safety remains a significant issue, with road crashes being a leading cause of fatalities and injuries. While several studies have been conducted on crash severity, few have analyzed and predicted specific types of crashes, such as fatal crashes. Identifying the key factors associated with fatal crashes and predicting their occurrence can help develop effective preventative measures. Objective This study intended to develop cluster analysis and ML-based models using crash data to extract the prominent factors behind fatal crash occurrences and analyze the inherent pattern of variables contributing to fatal crashes. Methods Several branches and categories of supervised ML models have been implemented for fatality prediction and their results have been compared. SHAP analysis was conducted using the ML model to explore the contributing factors of fatal crashes. Additionally, the underlying hidden patterns of fatal crashes have been evaluated using K-means clustering, and specific fatal crash scenarios have been extracted. Results The deep neural networks model achieved 85% accuracy in predicting fatal crashes in Kansas. Factors, such as speed limits, nighttime, darker road conditions, two-lane highways, highway interchange areas, motorcycle and tractor-trailer involvement, and head-on collisions were found to be influential. Moreover, the clusters were able to discern certain scenarios of fatal crashes. Conclusion The study can provide a clear image of the important factors related to fatal crashes, which can be utilized to create new safety protocols and countermeasures to reduce fatal crashes. The results from cluster analysis can facilitate transportation professionals with representative scenarios, which will benefit in identifying potential fatal crash conditions.

Study and methodology exploration of high-potassium glass and lead-barium glass classification patterns based on cluster analysis and decision tree

This study aims to analyze the classification patterns of high-potassium glass and lead-barium glass. It involves preprocessing the data through cluster analysis and employing machine learning decision trees and K-means clustering models for in-depth analysis. In the clustering analysis model, we start by performing a cluster difference analysis on the feature attributes of high-potassium glass and lead-barium glass to obtain initial classification results. Subsequently, we use machine learning decision tree models to partition the data into training and testing sets to explore the classification patterns. Construct a K-means clustering model by iteratively determining the initial cluster centroids' positions using an algorithm and calculating the silhouette coefficient for different numbers of clusters.This process aids in determining the appropriate number of clusters and the initial cluster centroids' positions. The results indicate significant differences in various features between high-potassium glass and lead-barium glass, and the employment of decision trees and K-means models successfully classifies artifacts. In conclusion, this study provides a robust method for the classification of high-potassium glass and lead-barium glass, offering valuable insights for research and application in related fields.