Agglomerative Hierarchical Clustering Research Articles

High vulnerability of coastal wetlands in Chile at multiple scales derived from climate change, urbanization, and exotic forest plantations

Coastal wetlands are considered one of the most vulnerable ecosystems worldwide; the ecosystem services they provide and the conservation of their biodiversity are threatened. Despite the high ecological and socioenvironmental value of coastal wetlands, regional and national vulnerability assessments are scarce. In this study we aimed to assess the vulnerability of coastal wetlands in Chile from 18°S to 42°S (n = 757) under a multiscale approach that included drivers associated with climate change and land cover change. We assessed multiple drivers of vulnerability at three spatial scales (10 m, 100 m, and 500 m) by analyzing multiple remote sensing data (16 variables) on land cover change, wildfires, climatic variables, vegetation functional properties, water surface and importance for biodiversity. We constructed a multifactorial vulnerability index based on the variables analyzed, which provided a map of coastal wetland vulnerability. Then we explored the main drivers associated with the vulnerability of each coastal wetland by performing a Principal Components Analysis with Agglomerative Hierarchical Clustering, which allowed us to group coastal wetlands according to the drivers analyzed. We found that 42.6 ± 9.2 % of the coastal wetlands evaluated have high or very high vulnerability, with higher vulnerability at the 500 m scale (51.4 %). We identified four groups of coastal wetlands: two located in central Chile, mainly affected by climate change-associated drivers (41.9 ± 2.1 %), and one in central Chile which is affected by land cover change (52.8 ± 6.2 %); the latter has a lower vulnerability level. The most vulnerable coastal wetlands were located in central Chile. Our results present novel findings about the current vulnerability of coastal wetlands, which could be validated by governmental institutions in field campaigns. Finally, we believe that our methodological approach could be useful to generate similar assessments in other world zones.

Who does not use urban green spaces and why? Insights from a comparative study of thirty-three European countries

There is a large body of research on the benefits green spaces offer city dwellers and how people use urban green spaces (UGS). However, there is much less information on how many people do not use UGS, who these people are, and why they resign from going there. This paper focuses on understanding reasons that restrict the use of UGS and draws the socio-economic and demographic profile of UGS non-users. For this purpose, we surveyed N = 10,391 respondents from 33 European countries (statistically representative at the country level). Results show that about 10% of the respondents do not use UGS. However, the share of non-users differs between countries from 2% (Turkey) to 25% (United Kingdom). Results reveal statistically significant differences between those who do not use and those who use UGS across age, education level, and residential location, demonstrating the importance of considering the needs of different groups of people when designing, planning and managing UGS. Lack of time and interest, and long distance from where people live are key limiting factors. The dominant reasons for not using UGS enable us to group the countries using the hierarchical agglomerative clustering analysis. With this study, we contribute to the knowledge of personal barriers that prevent UGS use and, in particular, connect the different reasons for not using UGS with the socio-economic groups typically listed as deprived when it comes to UGS provision.

Load profiling of commercial and residential building using clustering technique

Data mining is a promising tool used in processing energy data collected from energy consumers. The knowledge derived from energy data is very pertinent in the formulation of various demand-side management programs. This paper uses clustering techniques to segment the energy consumption patterns of residential and commercial buildings; situated at different geographical locations. The two (2) commonly used clustering techniques: K-Means and Agglomerative Hierarchical Clustering, were employed. The result indicates that the choice of clustering technique for load profiling is highly subjective to the nature of the dataset. Hence, using Davies-Bouldin Index (DB) Index and Silhouette Index (SI) as clustering indicators to select an optimum number of clusters and the best clustering technique. Hierarchical clustering was identified as the most appropriate clustering for the two buildings.

Cloud-Based Dynamic Task Scheduling and Load Balancing Using a Variety of Intelligent Schemes

One of the main problems with cloud computing is load balancing, which divides tasks into virtual machines (VMs) with varying lengths and completion times. As a result, keeping VMs loaded up during job scheduling and allocation is difficult. In this research, we suggested secure load balancing in a cloud context and resource-aware dynamic job scheduling utilizing intelligent algorithms. The process for the planned work is as follows: (1). The Human Mental Search (HMS) method is used to cluster Real-Time and Non-Real Time jobs into four categories: Real-Time and Short, Real-Time and Long, Non-Real Time and Short, and Non-Real Time and Long. (2). Knowledge of Energy and Delay Combined Tasks Tasks are planned based on task clustering, and the fuzzy VIKOR method is used to schedule them in the order of priority. (3). Security Risks aware VM Clustering, which employs the Hierarchical Agglomerative Clustering algorithm, consists of three distinct entities: a VM's information collector, which compiles data about VMs; a load monitor, which oversees the entire datacenter; and a decision maker, which assigns tasks by choosing the best VM. (4). Co-Resident assaults were stopped by the continuous-Actor-Critical-Algorithm (C- ACAL), which is used to pick the best VM. Using CloudSim, which measures performance in terms of response time, latency, resource utilization, throughput, energy consumption, and task execution success rate, the simulation is carried out.

Identification of volatile organic compounds (VOCs) by SPME-GC-MS to detect Aspergillus flavus infection in pistachios

Aflatoxin contamination poses a serious problem for the pistachio industry as aflatoxin B1 (AFB1), the most harmful aflatoxin, can be produced by some species of the Aspergillus and Penicillium. Like any other living organism, Aspergillus flavus (A. flavus) releases specific volatile organic compounds (VOCs) as a result of its metabolic reactions. In this study, we extracted and analyzed the VOCs produced by toxigenic A. flavus using solid-phase microextraction (SPME) in combination with gas chromatography-mass spectroscopy (GC-MS). The study was conducted on culture media and pistachio inoculated samples over a period of 16 days to observe the production of VOCs and aflatoxin over time. Various chemicals from different classes (such as organic acids, alcohols, ketones, and hydrocarbons) were detected during the growth of A. flavus on pistachios. Agglomerative hierarchical clustering was performed on the GC-MS spectra to identify VOCs corresponding to AFB1. We have identified ten VOCs with the potential to function as biomarkers for monitoring the growth of A. flavus NRRL 3357 on in-shell pistachios, and α,β-dimethylbenzenepropanoic acid was identified as the primary early marker for detecting A. flavus infection in pistachios.

Garment sizing system development for Amhara policemen uniforms using data mining techniques

Purpose Sizing system is a fundamental topic in garment fitting. The purpose of this study was to assess the fit of existing police uniforms (shirt, jacket, overcoat and trousers) and develop a sizing system for upper and lower body uniforms of Amhara policemen in Ethiopia. Design/methodology/approach In total, 35 body dimensions of 889 policemen were taken through a manual anthropometric survey following the procedures in ISO 8559:1989 after each subject was interviewed on issues related to garment fit. The anthropometric data were pre-processed, key body dimensions were identified by principal components analysis and body types were clustered by the agglomerative hierarchical clustering algorithm and verified by the XGBoost classifier in a Python programming environment. The developed size charts were validated statistically using aggregate loss and accommodation rate. Findings About 44% of the subjects encountered fit problems every time they own new readymade uniforms. Lengths and side seams of shirts, and lengths and waist girths of trousers are the most frequently altered garment sites. Analysis of the anthropometric measurements resulted in 13 and 15 sizes for the upper and lower bodies, respectively. Moreover, the comparison of the developed upper garment size chart with the existing size chart for a shirt showed a considerable difference. This indicates that inappropriate size charts create fit problems. Originality/value The study considers the analysis of fit problems and sizing system development in a less researched country. Moreover, the proposed data mining procedure and its application for size chart development is unique and workable.

Machine learning and statistical models for analyzing multilevel patent data

A recent surge of patent applications among public hospitals in China has aroused significant research interest. A country’s healthcare innovation capacity can be measured by its number of patents. This paper explores the link between the number of patents and ten independent variables. Multicollinearity was carefully detected and removed by using the variable selection method and LASSO regression, respectively. The Poisson model and the negative binomial model were proposed to analyze the patent data. Three goodness of fit tests, the Pearson test, the deviance test, and the DHARMa non-parametric dispersion test, were conducted to investigate if the model has a good fit. After discovering four clusters by conducting agglomerative hierarchical clustering, these two models were replaced by the negative binomial mixed model. The likelihood ratio test was used to determine which model is more appropriate and the results reveal that the negative binomial mixed model outperforms both the Poisson model and the negative binomial model. Three variables, number of health technicians per 10,000 people, financial expenditure on science and technology as well as number of patent applications per 10,000 health personnel, have a significantly positive relationship with the number of patents in Chinese tertiary public hospitals.

Dynamics of crop category choices reveal strategies and tactics used by smallholder farmers in India to cope with unreliable water availability

CONTEXTChanging cropping pattern is a potentially effective lever to cope with unreliable water resources, but given the multiple factors driving crop choices, assessing if farmers actually use it specifically for this objective remains difficult. OBJECTIVEWe aimed at analyzing whether and how farmers choose crop categories with different water requirements to cope with limited water resources in peninsular India. METHODSWe monitored over a 10 year period crop choices, weather, and groundwater level for the three cropping seasons in 205 irrigable farms in the Berambadi watershed, in southern India. We categorized crops according to their seasonal water requirement. We built farm types based of Sequence Analysis and Agglomerative Hierarchical Clustering of crop category sequences over 10 years for each cropping season. For each type, we correlated the variation in crop category choices to variations in rainfall and groundwater availability, to identify tactical adaptations. Finally, we grouped the crop category choices of the three seasons to identify the main strategic pathways followed by farmers. RESULTS AND CONCLUSIONSSequence analysis of crop category choices revealed different types of crop category sequences, reflecting farmers' different strategies, which were not significantly linked with groundwater availability. We identified five main pathways across the three cropping seasons, including combining long-cycle irrigated crop and other crop categories, specializing in short-cycle irrigated crops over two or three seasons, specializing in rainfed crops or abandoning agriculture. Within each type, correlations between variation in water availability and crop categories highlighted specific tactical adaptations. SIGNIFICANCEThe opportunity for farmers to choose their crops among a variety of species encompassing a large range of water requirements allows them to base their system resilience on a large diversity of strategies and tactics. This suggests that some farmers empirically estimate the water balance of their cropping systems at seasonal scale to take tactical decisions. Providing them with science-based tools to refine this estimation could therefore improve their decision-making. This also implies that modelling farmer decisions must account for their diversity. Maintaining or increasing the capacity of farmers to cultivate a broad range of crops with different seasonal water requirements is important for farming system resilience, and should therefore be part of the agenda of policy makers for agricultural or environmental regulations.

Sentinel-2 satellite images for monitoring cattle slurry and digestate spreading on emerging wheat crop: a field spectroscopy experiment

This study is aimed to evaluate the utility of Sentinel-2 imagery for monitoring exogenous organic matter (EOM) applied on winter wheat crop, using two spatial scales: proximal and satellite. From proximal sensing, multi-temporal spectral field measurements were taken on experimental fields consisting of three treatments (cattle slurry, liquid and raw digestates) and a control throughout 46 days. From Sentinel-2 satellites, images were analysed before and after EOM application. For both sensing scales, EOM and vegetation indices were used. On any scale of observation, the digestates spread on emerging wheat were easily detectable in late winter, in contrast to spring spreading events which were hindered by the developed vegetation. The agglomerative hierarchical clustering from the EOM indices divided by EVI achieved to discriminate digestates at early and medium stages of vegetation growth. Our findings did not apply for cattle slurry, presumably because of both lower organic and dry matter contents.

Antineutrophil cytoplasmic antibody-associated vasculitis classification by cluster analysis based on clinical phenotypes: a single-center retrospective cohort study.

Antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) refers to a group of small vessel inflammatory disorders. Overlapping clinical phenotypes of AAV subgroups continually provoke controversies over their diagnostic and classification criteria. Using the agglomerative hierarchical clustering method, we classified 210 Korean patients diagnosed with AAV into mutually exclusive clusters according to Birmingham Vasculitis Activity Score items, ANCA specificity, sex, and age. We analyzed the resulting clusters' outcomes to investigate the clinical significance of the classification. We proposed a distance-based algorithm of patient assignment and explored its clinically relevant modification. In total, 116 patients (55%) had microscopic polyangiitis, 53 (25%) had granulomatosis with polyangiitis, and 42 (20%) had eosinophilic granulomatosis with polyangiitis. Our model grouped the patients into five clusters, namely, "limited proteinase 3 (PR3)-ANCA vasculitis," "generalized PR3-ANCA vasculitis," "ANCA-negative vasculitis," "renal-limited vasculitis," and "myeloperoxidase-ANCA vasculitis." Patients clustered under "generalized PR3-ANCA vasculitis" had a higher relapse rate (hazard ratio [HR] = 2.12, P = 0.067). The incidence of end-stage renal disease was higher in patients belonging to the "renal-limited vasculitis" cluster (HR=1.50, P=0.03), and those in the "ANCA-negative vasculitis" cluster experienced a relatively milder clinical course of AAV (mortality = 0). Because the clusters were naturally derived from their distinguished phenotypes and have different clinical courses, our clustering method may be a more clinically relevant classification system for AAV, revealing its phenotypic diversity. We also proposed a simple and intuitive distance-based assignment algorithm, which can be easily modified according to specific clinical needs. Key Points • In this study with a single-center AAV cohort, we showed that AAV can be divided into five distinct subclasses with different disease courses based on the clinical and laboratory features of the patients. • Our study revealed ethnic differences in AAV manifestation and suggests that physicians may need to analyze their own AAV patients to assess the disease status of AAV patients. • We proposed a distance-based cluster membership assignment method that can be clinically modified to fit the specific purpose of grouping patients.

Characterising symptom clusters in patients with atrial fibrillation undergoing catheter ablation

ObjectiveThis study aims to leverage natural language processing (NLP) and machine learning clustering analyses to (1) identify co-occurring symptoms of patients undergoing catheter ablation for atrial fibrillation (AF) and (2)...

Data-Driven Characteristics of Patients Treated With Drug-Coated Balloon for De Novo Small Coronary Artery Disease: Hierarchical Agglomerative Clustering Analysis in the Registry

Data-Driven Characteristics of Patients Treated With Drug-Coated Balloon for De Novo Small Coronary Artery Disease: Hierarchical Agglomerative Clustering Analysis in the Registry

Morpho-molecular exploration and selection of elite genotypes from indigenous Syzygium cumini L. Skeels (jamun) diversity of North-Western Indian Himalayas

AbstractEfficiently distinguishing various Syzygium cumini L. Skeels (jamun) accessions holds practical significance for selection purposes. This study concentrated on 15 superior genotypes of jamun from the North Western Indian Himalayas, selected for their pivotal horticultural traits. Drawn from a pool of 82 collected genotypes and assessed across two consecutive years (2019 and 2020), these genotypes underwent morphological evaluations utilizing a randomized block design replicated thrice. Concurrently, random amplified polymorphic DNA (RAPD) and inter simple sequence repeat (ISSR) markers were employed for molecular analysis. Substantial variations surfaced among genotypes, both in morphological traits and fruit biochemistry. Notably, tree 43 exhibited promise across multiple horticultural facets, encompassing fruit weight, length, pulp weight, pulp-to-seed ratio and pulp percentage. Conversely, tree 49 excelled in elevated levels of total soluble solids, total sugar and reducing sugar. While principal component analysis and cluster analysis unveiled modest genetic variability, RAPD and ISSR markers unveiled pronounced molecular-level polymorphism. Agglomerative hierarchical clustering delineated the genotypes into five distinct clusters. Cluster I encompassed two genotypes, cluster II embraced five while the largest group, cluster III, included six genotypes. Clusters IV and V highlighted individual genotypes, trees 43 and 54 respectively. In the molecular analysis, UPGMA clustering yielded two primary clusters, spotlighting the noteworthy similarity between genotypes trees 49 and 52 whereas, trees 40, 43, 44 and 48 stood distinct. The observed genetic diversity stands as a valuable resource with substantial potential to enrich diverse breeding initiatives. These salient genetic variations underscore the richness within the studied population, offering a valuable asset for focused future pursuits.

Association of comorbid-socioeconomic clusters with mortality in late onset epilepsy derived through unsupervised machine learning

Background and objectivesLate-onset epilepsy is a heterogenous entity associated with specific aetiologies and an elevated risk of premature mortality. Specific multimorbid-socioeconomic profiles and their unique prognostic trajectories have not been described. We sought to determine if specific clusters of late onset epilepsy exist, and whether they have unique hazards of premature mortality. MethodsWe performed a retrospective observational cohort study linking primary and hospital-based UK electronic health records with vital statistics data (covering years 1998–2019) to identify all cases of incident late onset epilepsy (from people aged ≥65) and 1:10 age, sex, and GP practice-matched controls. We applied hierarchical agglomerative clustering using common aetiologies identified at baseline to define multimorbid-socioeconomic profiles, compare hazards of early mortality, and tabulating causes of death stratified by cluster. ResultsFrom 1,032,129 people aged ≥65, we identified 1048 cases of late onset epilepsy who were matched to 10,259 controls. Median age at epilepsy diagnosis was 68 (interquartile range: 66–72) and 474 (45%) were female. The hazard of premature mortality related to late-onset epilepsy was higher than matched controls (hazard ratio [HR] 1.73; 95% confidence interval [95%CI] 1.51–1.99). Ten unique phenotypic clusters were identified, defined by ‘healthy’ males and females, ischaemic stroke, intracerebral haemorrhage (ICH), ICH and alcohol misuse, dementia and anxiety, anxiety, depression in males and females, and brain tumours. Cluster-specific hazards were often similar to that derived for late-onset epilepsy as a whole. Clusters that differed significantly from the base late-onset epilepsy hazard were ‘dementia and anxiety’ (HR 5.36; 95%CI 3.31–8.68), ‘brain tumour’ (HR 4.97; 95%CI 2.89–8.56), ‘ICH and alcohol misuse’ (HR 2.91; 95%CI 1.76–4.81), and ‘ischaemic stroke’ (HR 2.83; 95%CI 1.83–4.04). These cluster-specific risks were also elevated compared to those derived for tumours, dementia, ischaemic stroke, and ICH in the whole population. Seizure-related cause of death was uncommon and restricted to the ICH, ICH and alcohol misuse, and healthy female clusters. SignificanceLate-onset epilepsy is an amalgam of unique phenotypic clusters that can be quantitatively defined. Late-onset epilepsy and cluster-specific comorbid profiles have complex effects on premature mortality above and beyond the base rates attributed to epilepsy and cluster-defining comorbidities alone.

Machine learning for lumbar and pelvis kinematics clustering

Clustering algorithms such as k-means and agglomerative hierarchical clustering (HCA) may provide a unique opportunity to analyze time-series kinematic data. Here we present an approach for determining number of clusters and which clustering algorithm to use on time-series lumbar and pelvis kinematic data. Cluster evaluation measures such as silhouette coefficient, elbow method, Dunn Index, and gap statistic were used to evaluate the quality of decision making. The result show that multiple clustering evaluation methods should be used to determine the ideal number of clusters and algorithm suitable for clustering time-series data for each dataset being analyzed.

Flavour profile, consumer acceptance and cooking methods of Portulacaria afra (spekboom) leaves

Portulacaria afra (P. afra) is a highly adaptable edible wild-growing succulent plant known as Spekboom. It thrives in gardens and the wild in arid and semi-arid regions of Southern Africa. It has a pleasant and citrusy acidic flavour. The study determined the consumer liking and sensory profile of five Spekboom treatments (raw, blanched, oven-baked, stir-fried and steamed). Ninety-four panellists ranked it on a nine-point hedonic scale and selected sensory characteristics which best described the taste using the Check-All-That-Apply (CATA) assessment. The overall liking assessment showed the fresh sample was ‘neither liked nor disliked’ (5.06), while the other treatments were ‘disliked slightly’. The most frequently checked terms in the CATA assessment were “leafy appearance” (201) and sour taste (369). Agglomerative Hierarchical Clustering and External Preference Mapping indicated two clusters. Cluster 1 indicated the raw leaves as fresh, green, leafy, and crispy, while the second cluster preferred blanched, oven-baked, and steamed leaves with descriptors like sour, bitter, chewy, juicy and pungent. All the treatments, except stir-frying, appeared to be acceptable to consumers. Thus, this plant has an acceptable taste. It could assist in ensuring sustainable food security and reduce the vulnerability of rural and urban food-insecure households in Southern Africa.

Use of unsupervised machine learning to characterise HIV predictors in sub-Saharan Africa

IntroductionSignificant regional variations in the HIV epidemic hurt effective common interventions in sub-Saharan Africa. It is crucial to analyze HIV positivity distributions within clusters and assess the homogeneity of countries. We aim at identifying clusters of countries based on socio-behavioural predictors of HIV for screening.MethodWe used an agglomerative hierarchical, unsupervised machine learning, approach for clustering to analyse data for 146,733 male and 155,622 female respondents from 13 sub-Saharan African countries with 20 and 26 features, respectively, using Population-based HIV Impact Assessment (PHIA) data from the survey years 2015–2019. We employed agglomerative hierarchical clustering and optimal silhouette index criterion to identify clusters of countries based on the similarity of socio-behavioural characteristics. We analyse the distribution of HIV positivity with socio-behavioural predictors of HIV within each cluster.ResultsTwo principal components were obtained, with the first describing 62.3% and 70.1% and the second explaining 18.3% and 20.6% variance of the total socio-behavioural variation in females and males, respectively. Two clusters per sex were identified, and the most predictor features in both sexes were: relationship with family head, enrolled in school, circumcision status for males, delayed pregnancy, work for payment in last 12 months, Urban area indicator, known HIV status and delayed pregnancy. The HIV positivity distribution with these variables was significant within each cluster.Conclusions /findingsThe findings provide a potential use of unsupervised machine learning approaches for substantially identifying clustered countries based on the underlying socio-behavioural characteristics.

A time series analysis of milk productivity in US dairy states

As US dairy cow production evolves, it is important to characterize trends and seasonal patterns to project amounts and fluctuations in milk and milk components by states or regions. Hence, this study aimed to (1) quantify historical trends and seasonal patterns of milk and milk components production associated with calving date by parities and states; (2) classify parities and states with similar trends and seasonal patterns into clusters; and (3) summarize the general pattern for each cluster for further application in simulation models. Our data set contained 9.18 million lactation records from 5.61 million Holstein cows distributed in 17 states during the period January 2006 to December 2016. Each record included a cow's total milk, fat, and protein yield during a lactation. We used time series decomposition to obtain each state's annual trend and seasonal pattern in milk productivity for each parity. Then, we classified states and parities with agglomerative hierarchical clustering into groups according to 2 methods: (1) dynamic time warping on the original time series and (2) Euclidean distance on extracted features of trend and seasonality from the decomposition. Results showed distinguishable trends and seasonality for all states and lactation numbers for all response variables. The clusters and cluster centroid pattern showed a general upward trend for all yields [energy-corrected milk (ECM), milk, fat, and protein] and a steady trend for fat and protein percent for all states except Texas. We also found a larger seasonality amplitude for all yields (ECM, milk, fat, and protein) from higher lactation numbers and a similar amplitude for fat and protein percent across lactation numbers. The results could be used for advising management decisions according to farm productivity goals. Furthermore, the trend and seasonality patterns could be used to adjust the production level in a specific state, year, and season for farm simulations to accurately project milk and milk components production.

Hierarchical agglomerative clustering-based skin lesion detection with region based neural networks classification

The skin cancer disease is widely spreading in the world due to abnormal increment of radiation. The segmentation of images is a crucial yet challenging aspect of the image processing process. It has emerged as a central focus in the study of image comprehension in recent years. Skin cancer is one of the critical diseases, which caused millions of deaths over the globe. The computer aided detection (CAD) of skin cancer can save the human lives. Further, the CAD models are resulted in poor segmentation and classification performance due to basic image processing, and machine learning models. Therefore, this work focused on development of skin lesion detection and classification (SLDC) using deep learning models. Initially, Guided image filtering (GIF) based preprocessing method, which performs noise removal, skin lesion enhancements, and basic hair removal operations. Then, skin lesion disease effected region is identified by using hierarchical agglomerative clustering (HAC). Then, multi-level features are extracted from segmented images. Finally, the region based neural network with long short term memory (RNN-LSTM) are used to perform the benign and malignant classes classification using trained features. The simulations conducted on International Symposium on Biomedical Imaging (ISBI) dataset shows that, the proposed RNN-SLDC system resulted in superior subjective and objective performance as compared to other approaches. The proposed method achieved 99.71% of segmentation accuracy, and 99.46% of classification accuracy.

Groundwater monitoring and specific yield estimation using time-lapse electrical resistivity imaging and machine learning

This paper presents an alternative method for monitoring groundwater levels and estimating specific yields of an unconfined aquifer under different seasonal conditions. The approach employs the Time-Lapse Electrical Resistivity Imaging (TL-ERI) method and machine learning-based time series clustering. A TL-ERI survey was conducted at ten sites (WS01-WS10 sites) throughout the dry and wet seasons, with five-time measurements collected for each site, in the Taichung-Nantou Basin along the Wu River, Central Taiwan. The obtained resistivity raw data was inverted and converted into normalized water content values using Archie’s law, followed by applying the Van Genuchten (VG) model for the Soil Water Characteristic Curve to estimate the Groundwater Level (GWL), and estimated the theoretical specific yield (Sy) by computing the difference between the saturated and residual water contents of the fitted VG model. In addition, the specific yield capacity (Sc), representing the nature of the storage capacity in the aquifer, was also calculated. The results showed that this approach was able to estimate those hydrogeological parameters. The spatial distribution of the GWL reveals that during the dry-wet seasons from February to July, there was a high GWL that extended from southeast to northwest. Conversely, during the wet-dry seasons from July to October, the high GWL shrank, which can be attributed to recharge variations from rainfall events. The determined spatial distribution of Sy and Sc fall within the range of 0.03–0.24 and 0.14–0.25, respectively. To quantitatively establish areas of similar groundwater level changes along with the VG model parameter variations during the study period, a Time series Clustering analysis (TSC) was performed by utilizing Hierarchical Agglomerative Clustering (HAC). The findings suggest that the WS03 site is a promising area for further investigation due to its highest Sc value with a slight change in groundwater levels during the dry and wet seasons. This study brings an advanced development of the geoelectrical method to estimate regional hydrogeological parameters in an area with limited available groundwater observation wells, in different seasonal conditions for groundwater management purposes.