Number Of Clusters Research Articles

Signal Intensity Trajectories Clustering for Liver Vasculature Segmentation and Labeling (LiVaS) on Contrast-Enhanced MR Images: A Feasibility Pilot Study.

This study aims to develop a semiautomated pipeline and user interface (LiVaS) for rapid segmentation and labeling of MRI liver vasculature and evaluate its time efficiency and accuracy against manual reference standard. Retrospective feasibility pilot study. Liver MR images from different scanners from 36 patients were included, and 4/36 patients were randomly selected for manual segmentation as referenced standard. The liver was segmented in each contrast phase and masks registered to the pre-contrast segmentation. Voxel-wise signal trajectories were clustered using the k-means algorithm. Voxel clusters that best segment the liver vessels were selected and labeled by three independent radiologists and a research scientist using LiVaS. Segmentation times were compared using a paired-sample t-test on log-transformed data. The agreement was analyzed qualitatively and quantitatively using DSC for hepatic and portal vein segmentations. The mean segmentation time among four readers was significantly shorter than manual (3.6 ± 1.4 vs. 70.0 ± 29.2min; p < 0.001), even when using a higher number of clusters to enhance accuracy. The DSC for portal and hepatic veins reached up to 0.69 and 0.70, respectively. LiVaS segmentations were overall of good quality, with variations in performance related to the presence/severity of liver disease, acquisition timing, and image quality. Our semi-automated pipeline was robust to different MRI vendors in producing segmentation and labeling of liver vasculature in agreement with expert manual annotations, with significantly higher time efficiency. LiVaS could facilitate the creation of large, annotated datasets for training and validation of neural networks for automated MRI liver vascularity segmentation. HIGHLIGHTS: Key Finding: In this pilot feasibility study, our semiautomated pipeline for segmentation of liver vascularity (LiVaS) on MR images produced segmentations with simultaneous labeling of portal and hepatic veins in good agreement with the manual reference standard but at significantly shorter times (mean LiVaS 3.6 ± 1.4 vs. mean manual 70.0 ± 29.2min; p < 0.001). Importance: LiVaS was robust in producing liver MRI vascular segmentations across images from different scanners in agreement with expert manual annotations, with significant ly higher time efficiency, and therefore potential scalability.

An Exploratory Analysis of Physical Examination Subtypes in Pediatric Athletes With Concussion.

Pediatric athletes with concussion present with a variety of impairments on clinical assessment and require individualized treatment. The Buffalo Concussion Physical Examination is a brief, pertinent clinical assessment for individuals with concussion. The purpose of this study was to identify physical examination subtypes in pediatric athletes with concussion within 2 weeks of injury that are relevant to diagnosis and treatment. Secondary analysis of a published cohort study and clinician consensus. Three university-affiliated sports medicine centers. Two hundred seventy children (14.9 ± 1.9 years). Orthostatic intolerance, horizontal and vertical saccades, smooth pursuits, vestibulo-ocular reflex, near-point convergence, complex tandem gait, neck range of motion, neck tenderness, and neck spasm. Correlations between independent variables were calculated, and network graphs were made. k-means and hierarchical clustering were used to identify clusters of impairments. Optimal number of clusters was assessed. Results were reviewed by experienced clinicians and consensus was reached on proposed subtypes. Physical examination clusters overlapped with each other, and no optimal number of clusters was identified. Clinician consensus suggested 3 possible subtypes: (1) visio-vestibular (horizontal and vertical saccades, smooth pursuits, and vestibulo-ocular reflex), (2) cervicogenic (neck range of motion and spasm), and (3) autonomic/balance (orthostatic intolerance and complex tandem gait). Although we identified 3 physical examination subtypes, it seemed that physical examination findings alone are not enough to define subtypes that are both statistically supported and clinically relevant, likely because they do not include symptoms, assessment of mood or cognitive problems, or graded exertion testing.

The Importance of Digital Technology and Clustering for Innovation in MSEs. Evidence from Secondary Data in Indonesia

This paper explores the relationship between innovation and these two determinants in micro and small enterprises (MSEs) in the manufacturing industry in Indonesia, using secondary data. The key question in this research: is there evidence of the importance of DT and clustering for innovation capability in MSEs in Indonesia? The paper analyzes secondary data from Indonesia’s National Agency of Statistics and reviews key literature on innovation and digitalization in MSEs, and industrial cluster development in Indonesia. It is found that the number of industrial clusters of micro-, small-, and medium-sized enterprises (MSMEs) is concentrated on the island of Java, where more than 50% of the population is located, and is the most advanced region in economic development and industrialization in Indonesia. The majority of MSME clusters are in the food industry and woodworking industries. Only a very few of MSEs in the manufacturing industry use the Internet and do innovation. The type of innovation mostly is product innovation. Both relationships between MSEs doing innovation and MSEs using the internet and the total number of clusters are positive. But, statistically, only the relationship between the number of MSEs doing innovation and use of the internet is significant.

DendroX: multi-level multi-cluster selection in dendrograms

BackgroundCluster heatmaps are widely used in biology and other fields to uncover clustering patterns in data matrices. Most cluster heatmap packages provide utility functions to divide the dendrograms at a certain level to obtain clusters, but it is often difficult to locate the appropriate cut in the dendrogram to obtain the clusters seen in the heatmap or computed by a statistical method. Multiple cuts are required if the clusters locate at different levels in the dendrogram.ResultsWe developed DendroX, a web app that provides interactive visualization of a dendrogram where users can divide the dendrogram at any level and in any number of clusters and pass the labels of the identified clusters for functional analysis. Helper functions are provided to extract linkage matrices from cluster heatmap objects in R or Python to serve as input to the app. A graphic user interface was also developed to help prepare input files for DendroX from data matrices stored in delimited text files. The app is scalable and has been tested on dendrograms with tens of thousands of leaf nodes. As a case study, we clustered the gene expression signatures of 297 bioactive chemical compounds in the LINCS L1000 dataset and visualized them in DendroX. Seventeen biologically meaningful clusters were identified based on the structure of the dendrogram and the expression patterns in the heatmap. We found that one of the clusters consisting of mostly naturally occurring compounds is not previously reported and has its members sharing broad anticancer, anti-inflammatory and antioxidant activities.ConclusionsDendroX solves the problem of matching visually and computationally determined clusters in a cluster heatmap and helps users navigate among different parts of a dendrogram. The identification of a cluster of naturally occurring compounds with shared bioactivities implicates a convergence of biological effects through divergent mechanisms.

Finding repeatable progressive pass clusters and application in international football

Progressive passing in football (soccer) is a key aspect in creating positive possession outcomes. Whilst this is well established, there is not a consistent way to describe the different types of progressive passes. We expand on the previous literature, providing a complete methodological approach to progressive pass clustering from selection of the number of clusters (k) to risk-reward profiling of these progressive pass types. In this paper the Separation and Concordance (SeCo) framework is utilised to provide a process to analyse k-means clustering solutions in a more repeatable way. The results demonstrate that we can find stable progressive pass clusters in International Football and their efficacy with progressive passes “Mid Central to Mid Half Space” in build-up and “Mid Half Space to Final Central” into the final 3rd having the best balance between risk (turnover) and reward (shot created) in the subsequent possession. This allowed for opposition profiling of player and team patterns in different phases of play, with a case study presented for the teams in the Last 16 of the 2022 World Cup.

The response of yield, number of clusters, and cluster weight to meteorological factors and irrigation practices in grapevines: A multi-experiment study

Knowledge of the yield components of wine grapevines is essential to achieve target production and effectively design vineyard management. Therefore, it is necessary to map and analyze the factors that control the yield components to optimize wine production and profitability. The objective of this study was to identify the ranked effects of environmental and agrotechnical factors (e.g., meteorological, physiological, vegetative, and irrigation) on grapevine yield, number of clusters, and cluster weight, and to quantify their patterns. To achieve this objective, a multi-experiment database was used, assembling data collected during four long-term irrigation experiments including different varieties and diverse climatic regions across Israel. First, a paired analysis of the yield components was conducted to define the relationships among them. Then, prediction models for the three variables (i.e., yield components) were designed using the support vector machine (SVM) algorithm to explore the ranked effects and patterns of the controlling factors. Validation was conducted using a test set, consisting of 20% of the dataset. The relationships between yield and cluster weight (0 < r < 0.73) and between yield and the number of clusters were always positive (0.67 < r < 0.77), while negative relationships were found between cluster weight and number of clusters (-0.61 < r < 0). The accuracy of the prediction models was high, with mean absolute errors (normalized to the range) ranging between 4.1%-8.11%. The most dominant factors affecting yield components were cumulative leaf area index, accumulated growing degree days (GDD), cumulative irrigation, and evapotranspiration. Leaf area showed an optimum point, below and above which the values of the yield components decrease; GDD negatively affected yield and cluster weight but had a positive effect on the number of clusters; irrigation was positively associated with yield components but leveled out above 348-395 mm per season. Reference evapotranspiration had a negative effect on yield and cluster weight, and a generally positive effect on the number of clusters. Utilizing data from multiple experiments and applying multivariate analysis, allow to define general patterns of relationships and interactions between controlling factors and yield components. Extracting this knowledge from the data and managing the growing amounts of information can be achieved through complex modeling and the implementation of standard protocols for data collection and monitoring.

DRBM-ClustNet: A Deep Restricted Boltzmann-Kohonen Architecture for Data Clustering.

A Bayesian deep restricted Boltzmann-Kohonen architecture for data clustering termed deep restricted Boltzmann machine (DRBM)-ClustNet is proposed. This core-clustering engine consists of a DRBM for processing unlabeled data by creating new features that are uncorrelated and have large variance with each other. Next, the number of clusters is predicted using the Bayesian information criterion (BIC), followed by a Kohonen network (KN)-based clustering layer. The processing of unlabeled data is done in three stages for efficient clustering of the nonlinearly separable datasets. In the first stage, DRBM performs nonlinear feature extraction by capturing the highly complex data representation by projecting the feature vectors of d dimensions into n dimensions. Most clustering algorithms require the number of clusters to be decided a priori; hence, here, to automate the number of clusters in the second stage, we use BIC. In the third stage, the number of clusters derived from BIC forms the input for the KN, which performs clustering of the feature-extracted data obtained from the DRBM. This method overcomes the general disadvantages of clustering algorithms, such as the prior specification of the number of clusters, convergence to local optima, and poor clustering accuracy on nonlinear datasets. In this research, we use two synthetic datasets, 15 benchmark datasets from the UCI Machine Learning repository, and four image datasets to analyze the DRBM-ClustNet. The proposed framework is evaluated based on clustering accuracy and ranked against other state-of-the-art clustering methods. The obtained results demonstrate that the DRBM-ClustNet outperforms state-of-the-art clustering algorithms.

Adaptation rules inducing synchronization of heterogeneous Kuramoto oscillator network with triadic couplings.

A class of adaptation functions is found for which a synchronous mode with different number of phase clusters exists in a network of phase oscillators with triadic couplings. This mode is implemented in a fairly wide range of initial conditions and the maximum number of phase clusters is four. The joint influence of coupling strength and adaptation parameters on synchronization in the network has been studied. The desynchronization transition under variation of the adaptation parameter occurs abruptly and begins with the highest-frequency oscillator, spreading hierarchically to all other elements.

CLUSTERING DATA METEOROLOGI WILAYAH INDONESIA TIMUR DENGAN METODE K-MEANS DAN FUZZY C-MEANS

Climate change is a global issue that affect human life and the environment. Signs of climate change can be observed from long-term meteorological data. This research uses clustering techniques with the K-Means and Fuzzy C-Means methods to group cities in the Eastern Indonesia region based on numerical daily time series meteorological data from 1 January 2010 to 31 August 2023. The variables are minimum temperature, maximum temperature, temperature average, humidity, rainfall, duration of sunlight, maximum wind speed, and average wind speed. The dataset was collected from 28 meteorological stations. The K-Means and Fuzzy C-Means methods obtained the same results, namely the highest silhouette value of 0.218 with the number of clusters k = 2. In general, the annual trend shows an increase in temperature and a decrease in wind speed which are signs of climate change. This research is an early study of climate change in East Indonesia. The results of this research are expected to contribute to the study of climate change in Indonesia.

Effects of group size on movement patterns and clustering dynamics in rats.

Environment is a determining factor that can facilitate or hinder social interactions. A precursor to meaningfully engaging with conspecifics is being exposed to opportunistic encounters with others. Increasing the number of individuals in a given space (thus increasing density) would, statistically speaking, increase the likelihood of accidental encounters. This might have consequences on the formation of social networks-an idea that has not reliably been explored. If true, we would expect that increasing density would lead to an increase in the number and the duration of 'clusters' of animals. Here, we examined whether varying the number of rats in an open field environment differentially affected their movement dynamics or their propensity to aggregate into clusters and, if so, whether such effects are dependent solely on statistical factors due to increases in density, the potential for actively-sought social interactions, or both. We found that the number of rats in an environment impacts ambulation speed, distance traveled, cluster formation and approaches, and that number and duration of clusters are highly dependent on the propensity for the rats to engage in social interactions.

Improving the Quality Indicators of Multilevel Data Sampling Processing Models Based on Unsupervised Clustering

This paper presents a solution for building and implementing data processing models and experimentally evaluates new possibilities for improving ensemble methods based on multilevel data processing models. This study proposes a model to reduce the cost of retraining models when transforming data properties. The research objective is to improve the quality indicators of machine learning models when solving classification problems. The novelty is a method that uses a multilevel architecture of data processing models to determine the current data properties in segments at different levels and assign algorithms with the best quality indicators. This method differs from the known ones by using several model levels that analyze data properties and assign the best models to individual segments of data and training. The improvement consists of using unsupervised clustering of data samples. The resulting clusters are separate subsamples for assigning the best machine-learning models and algorithms. Experimental values of quality indicators for different classifiers on the whole sample and different segments were obtained. The findings show that unsupervised clustering using multilevel models can significantly improve the quality indicators of “weak” classifiers. The quality indicators of individual classifiers improve when the number of data clusters is increased to a certain threshold. The results obtained are applicable to classification when developing models and machine learning methods. The proposed method improved the classification quality indicators by 2–9% due to segmentation and the assignment of models with the best quality indicators in individual segments. Doi: 10.28991/ESJ-2024-08-01-025 Full Text: PDF

Cluster dynamics of proton irradiated RPV steel at room temperature

The life of reactor pressure vessels in nuclear power plants will be extended to 60 years, while the irradiation data of materials from commercial and experimental reactors are rare and expensive. There is an urgent need to solve how to accurately predict the changes in microstructure and mechanical properties of high-dose irradiated materials of RPV. The cluster dynamics method can be used to establish the relationship between nanostructure and macro properties and to study its evolution over a long period of years. Based on the average rate field theory, a cluster dynamics model was developed for the evolution of irradiated microstructure of RPV steel, which considers both homogeneous and heterogeneous nucleation of solute clusters. The model was successfully used to simulate the influence of the proton irradiation-induced cluster size and number density of RPV model steel with a low copper content at room temperature.

Functional subtypes of synaptic dynamics in mouse and human

Synapses preferentially respond to particular temporal patterns of activity with a large degree of heterogeneity that is informally or tacitly separated into classes. Yet, the precise number and properties of such classes are unclear. Do they exist on a continuum and, if so, when is it appropriate to divide that continuum into functional regions? In a large dataset of glutamatergic cortical connections, we perform model-based characterization to infer the number and characteristics of functionally distinct subtypes of synaptic dynamics. In rodent data, we find five clusters that partially converge with transgenic-associated subtypes. Strikingly, the application of the same clustering method in human data infers a highly similar number of clusters, supportive of stable clustering. This nuanced dictionary of functional subtypes shapes the heterogeneity of cortical synaptic dynamics and provides a lens into the basic motifs of information transmission in the brain.

Research on load curve fitting method based on K-means algorithm

Some areas of China are affected by the fluctuation of new energy output and load, and the power supply as a whole presents an unbalanced phenomenon of abundant electricity supply but shortage of power supply. To understand the characteristics of the load electricity consumption curve more accurately, this paper proposes a method of fitting the load electricity consumption curve based on K-mea ns algorithm. First, analyze the characteristics of the load electricity consumption based on the industry subdivision, and build the load electricity consumption characteristic index system. Then, the k-mea ns cluster analysis is carried out based on the characteristics of the load power consumption curve of the subdivided industries to determine the number of clusters, and the curve fitting and deviation test are carried out for each cluster group. This paper carries out practical analysis and theoretical research based on the data of typical industry load power consumption curve in Ningxia.

Pengelompokan Provinsi di Indonesia Menggunakan Time Series Clustering pada Sektor Ekspor Nonmigas

Indonesia's export activities are dominated by non-oil and gas exports consisting of four sectors, namely the processing industry, agriculture, mining, and others. The government must pay attention to non-oil and gas exports for each province because exports can play an essential role in a country's economic growth. This study was conducted to cluster provinces in Indonesia using time series clustering in the non-oil and gas export sector based on data patterns concerning Dynamic Time Warping (DTW) distance. The sectors used in this study are the manufacturing industry sector and the agricultural sector in 34 Indonesian provinces in the period 2017 - 2021. Time series clustering analysis uses the average linkage method with DTW distance and the selection of the optimum number of clusters using the silhouette coefficient method. The results of the analysis in the processing industry sector resulted in 3 optimum clusters, namely cluster 1 consisting of 1 province that has high processing industry exports, cluster 2 consisting of 8 provinces that have medium processing industry exports, and cluster 3 consisting of 25 provinces that have low processing industry exports. As for the agricultural sector, it produces 2 optimum clusters, namely cluster 1 consisting of 5 provinces that have high agricultural industry exports, and cluster 2 consisting of 29 provinces that have low agricultural industry exports. The clustering results in the processing industry sector and the agricultural sectors have a silhouette coefficient value of 0.778 and 0.798, so it is said to have a strong cluster structure.

A Theoretical Analysis of Density Peaks Clustering and the Component-Wise Peak-Finding Algorithm.

Density peaks clustering detects modes as points with high density and large distance to points of higher density. Each non-mode point is assigned to the same cluster as its nearest neighbor of higher density. Density peaks clustering has proved capable in applications, yet little work has been done to understand its theoretical properties or the characteristics of the clusterings it produces. Here, we prove that it consistently estimates the modes of the underlying density and correctly clusters the data with high probability. However, noise in the density estimates can lead to erroneous modes and incoherent cluster assignments. A novel clustering algorithm, Component-wise Peak-Finding (CPF), is proposed to remedy these issues. The improvements are twofold: 1) the assignment methodology is improved by applying the density peaks methodology within level sets of the estimated density; 2) the algorithm is not affected by spurious maxima of the density and hence is competent at automatically deciding the correct number of clusters. We present novel theoretical results, proving the consistency of CPF, as well as extensive experimental results demonstrating its exceptional performance. Finally, a semi-supervised version of CPF is presented, integrating clustering constraints to achieve excellent performance for an important problem in computer vision.

A comprehensive evaluation of OPTICS, GMM and K-means clustering methodologies for geochemical anomaly detection connected with sample catchment basins

The process of data-driven clustering to uncover geochemical anomalies linked to sample catchment basins (SCBs) includes a comprehensive framework to discern areas exhibiting unique geochemical attributes within a specified study area. The Ordering Points to Identify the Clustering Structure (OPTICS) method can serve as a robust methodology for detecting geochemical anomalies in SCBs. This is attributed to its capacity to effectively manage varying cluster densities, adaptively identify cluster numbers, exhibit resilience to noise, and display minimum sensitivity to parameters. A comparison was conducted in this research between the outcomes of the OPTICS clustering algorithm and two traditional clustering techniques, namely the Gaussian Mixture Model (GMM) and K-means clustering. In the following, the Expectation-Maximization (EM) technique is employed to train the GMM for clustering. Moreover, the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) as two validate statistical metrics implemented to ascertain the optimal number of components (clusters) belong to the GMM. It should be noted that the effectiveness of the clustering algorithms was further assessed using the Calinski-Harabasz (CH) index and the success-rate curves. OPTICS, a density-based clustering approach, was confirmed to be more effective than K-means and GMM for identifying MVT PbZn anomalies in Varcheh district, western Iran. Furthermore, the specified anomalies show a geo-spatial correspondence with the geological facts, and it has been observed that strong anomalies are more discoverable in close proximity to MVT PbZn occurrences. This work suggests a novel anomaly detection approach based on OPTICS, which exhibits superior performance and data-modeling efficiency. The main emphasis is on effectively distinguishing geochemical anomalies from sample data originating from populations with uncertain distributions.

The cosmological analysis of X-ray cluster surveys

Context. Cosmological studies have now entered Stage IV according to the Dark Energy Task Force (DETF) prescription. New missions (Euclid, Rubin Observatory, SRG/eROSITA) will cover very large fractions of the sky with unprecedented depth. These are expected to provide the required ultimate accuracy in the dark energy (DE) equation of state (EoS), which is required for the elucidation of the origin of the acceleration of cosmic expansion. However, none of these projects have the power to systematically unveil the galaxy cluster population in the 1 &lt; z &lt; 2 range. There therefore remains the need for an Athena-like mission to run independent cosmological investigations and scrutinise the consistency between the results from the 0 &lt; z &lt; 1 and 1 &lt; z &lt; 2 epochs. Aims. We study the constraints on the DE EoS and on primordial non-gaussanities for typical X-ray cluster surveys executed by a generic Athena-like Wide Field Imager. We focus on the impact of cluster number counts in the 1 &lt; z &lt; 2 range. Methods. We consider two survey designs: 50 deg2 at 80 ks (survey A) and 200 deg2 at 20 ks (survey B). We analytically derive cluster number counts and predict the cosmological potential of the corresponding samples, A and B, by means of a Fisher analysis. We adopt an approach that forward models the observed properties of the cluster population in the redshift–count rate–hardness ratio parameter space. Results. The achieved depth allows us to unveil the halo mass function down to the group scale out to z = 2. We predict the detection of thousands of clusters down to a few 1013h−1 M⊙, in particular 940 and 1400 clusters for surveys A and B, respectively, at z &gt; 1. Such samples will allow a detailed modelling of the evolution of cluster physics along with a standalone cosmological analysis. Our results suggest that survey B has the optimal design as it provides greater statistics. Remarkably, high-redshift clusters represent 15% or less of the full samples but contribute at a much higher level to the cosmological accuracy: by alleviating various degeneracies, these objects allow a significant reduction of the uncertainty on the cosmological parameters: Δwa is reduced by a factor of ∼2.3 and Δ fNLloc by a factor of ∼3. Conclusions. Inventorying the deep high-z X-ray cluster population can play a crucial role in ensuring overall cosmological consistency. This will be the major aim of future new-generation Athena-like missions.

Optimized Cluster-Based Cooperative Spectrum Sensing Over Weibull, Nakagami, Rician, and Rayleigh Fading Channels

This paper reports cluster-based cooperative spectrum sensing (CB-CSS) over Weibull, Nakagami, Rician, and Rayleigh fading channels in cognitive radio networks (CRNs). The performance assessment of centralized cooperative spectrum sensing (CCSS) and CB-CSS with different fusion rules over the Weibull fading channel is presented. The goal is to improve detection performance in a practical and efficient manner for CB-CSS. For this purpose, an expression for the optimal inter-fusion rule threshold (fusion rule between cluster head (CH) and fusion center (FC)) is proposed to minimize total error probability (TEP) over Weibull for CB-CSS. Simulations are conducted using analytical equations for missed detection and false alarm probability. The obtained expression for optimal inter-fusion rule threshold which minimizes TEP is verified by the results. The proposed model is validated with other fading channels such as Nakagami, Rician, and Rayleigh to show the superiority of the approach. It is observed that for a given range of detection threshold and for 4 × 6 cluster size (4 clusters and 6 CR users inside each cluster), the lowest total error probability is achieved over the Weibull fading channel. Minimum TEP over the Weibull fading channel for given network parameters is TEP = 1.656 × 10 − 6 when intra- and inter-fusion rules are l = 1 (OR fusion) and c = 3 . The least number of clusters and users essential to achieving target error bound over different fading channels are also determined.

PENGELOMPOKKAN KABUPATEN DI PROVINSI JAWA TENGAH BERDASARKAN KARAKTERISTIK IKLIM MENGGUNAKAN FUZZY CLUSTERING

There are many factors affecting human life, one of which is climate. Differences in climatic conditions in each region result in differences in the environment in society. The differences are referred to as potential natural resources, livelihoods, and social cultural conditions. The climate has an impact on culture in terms of how people dress, the shape of houses, and so on. The regency group in Central Java Province is based on similarities in climate characteristics using fuzzy clustering. The data used were taken from Central Java Provincial Statistical Office in 2022.The results of district grouping in Central Java Province are based on similarities in climate characteristics using fuzzy clustering with 4 different number of clusters and the validity tests of the Partition Coefficient and Classification Entropy indices. Based on the results of the index validity test, the optimal grouping results are 2 clusters with a Partition Coefficient value of 0.911233 and a Classification Entropy value of 0.07979. The 1st cluster consists of 13 districts with a cluster center at 27.9°C for air temperature, 2418mm for rainfall, and 804% for keelThe 2nd cluster and airbase consist of 22 cluster central districts at 26.6 °C for air temperature, 4087 mm for rainfall, and 81.3% for humidity.