Agglomerative Hierarchical Clustering Technique Research Articles

Delineating Precipitation Regions of the Contiguous United States from Cluster Analyzed Gridded Data

Spatially homogenous precipitation regions were delineated for the contiguous United States using a gridded data set of daily precipitation. Seasonal means (1981–2010) of four variables, together characterizing seasonal precipitation, were computed and subjected to a principal component analysis (PCA). PCA reduced the original 30,665 grid cells by sixteen precipitation variables (four variables, four seasons) in the data set. The standardized scores of the three retained principal components, which together retain 78.4 percent of the original data set’s variance, were then subjected to three agglomerative hierarchical clustering techniques. Using an objective method, several cluster solutions were examined, and the average linkage thirteen-cluster solution was deemed optimal. The average linkage solution was then subjected to a k-means partitioning technique under the premise that objects are not considered for reassignment during agglomerative hierarchical cluster procedures. The result is fifteen precipitation regions across the contiguous United States. Results indicate that the regions successfully minimize intraregion variability and maximize interregion variability when compared to the nine climate regions defined by the United States National Centers for Environmental Information. It is therefore suggested that the regions defined by this work will better serve research aimed at an improved understanding of long-term hydroclimate change and variability at regional to synoptic scales across the United States.

DELINEATION OF SITE-SPECIFIC MANAGEMENT ZONES USING MULTIVARIATE ANALYSIS AND GEOGRAPHIC INFORMATION SYSTEM TECHNIQUE

The basic objectives of this study were: 1) The ability to characterize the spatial variability across a soil for selected soil properties using GIS technique and 2) identification of site-specific management zones using selected soil properties using PCA and cluster analysis. 120 geo-referenced representative soil samples obtained from Sahl AlHussainiyah, El-Sharkia Governorate, Egypt, (from 0 to 0.60 m depth). These samples were prepared and analysed for soil characteristics, such as soil pH, electrical conductivity (ECe), calcium carbonate (CaCO3 ), soil organic matter (OM), available N, P, K, soil cations exchange capacity (CEC), and bulk density (BD). Using semi-variogram analysis and ordinary kriging, spatial distribution pattern varies from moderate to strong spatial dependence for most soil characteristics. Using PCA and cluster analysis, site-specific management zones were identified in the study area. For further analysis, four PCs with eigenvalues > 1 were used, with PCs explaining 73.19 percent of the variance. Four MZs were defined based on cluster analysis using an agglomerative hierarchical clustering technique. The differences between these MZs were statistically significant (p < 0.05). 3.01 (466.56 ha), 36.47 (5658.9 ha), 31.02 (4813.54 ha), and 29.5 percent (4577.8 ha) of the total area is MZ1, MZ2, MZ3, and MZ4 (15516.8 ha).

A Hybrid Model to Classify Patients with Chronic Obstructive Respiratory Diseases.

Over the last decades, an increase in the ageing population and age-related diseases has been observed, with the increase in healthcare costs. As so, new solutions to provide more efficient and affordable support to this group of patients are needed. Such solutions should never discard the user and instead should focus on promoting more healthy lifestyles and provide tools for patients’ active participation in the treatment and management of their diseases. In this concern, the Personal Health Empowerment (PHE) project presented in this paper aims to empower patients to monitor and improve their health, using personal data and technology assisted coaching. The work described in this paper focuses on defining an approach for user modelling on patients with chronic obstructive respiratory diseases using a hybrid modelling approach to identify different groups of users. A classification model with 90.4% prediction accuracy was generated combining agglomerative hierarchical clustering and decision tree classification techniques. Furthermore, this model identified 5 clusters which describe characteristics of 5 different types of users according to 7 generated rules. With the modelling approach defined in this study, a personalized coaching solution will be built considering patients with different necessities and capabilities and adapting the support provided, enabling the recognition of early signs of exacerbations and objective self-monitoring and treatment of the disease. The novel factor of this approach resides in the possibility to integrate personalized coaching technologies adapted to each kind of user within a smartphone-based application resulting in a reliable and affordable alternative for patients to manage their disease.

UAV-Thermal imaging and agglomerative hierarchical clustering techniques to evaluate and rank physiological performance of wheat genotypes on sodic soil

Sodicity is a major soil constraint in many arid and semi-arid regions worldwide, including Australia, which adversely affects the ability of crops to take up water and nutrients from the soil, reducing yield. Reliable methods and tools are required for appropriate selection of traits, may provide a better understanding of crop responses to multiple stresses, especially in sodic soil. A novel strategy was developed using unmanned aerial vehicle (UAV)-thermal imaging and agglomerative hierarchical clustering-based techniques to evaluate and rank the physiological performance of 18 contrasting wheat genotypes grown on a moderately sodic and a highly sodic soil in north-eastern Australia. We obtained UAV-thermal imaging data at different times of the day (9:30, 12:00, and 15:00 hrs) close to flowering stage. Crop biophysical parameters (Leaf potassium concentration, normalized difference vegetation index, crop water uptake, stomatal conductance, plant moisture content, and aboveground biomass) were measured at close to flowering by destructive plant sampling and ground-based proximal sensing and yield was machine harvested at maturity. Canopy temperatures derived from thermal imagery between 28.9 and 35.4 °C were observed at the moderately sodic site, and between 36.2 and 41.0 °C at the highly sodic site from 9:30 to 15:00 hrs. Canopy temperature was consistently higher than corresponding ambient air temperatures indicating plant water stress at both sites. While the air temperature was not significantly different (p > 0.05) between the two sites, canopy temperature was significantly higher (p < 0.01) on highly sodic soil compared to moderately sodic soil, indicating greater water stress at the highly sodic site. This difference was most likely due to the adverse impacts of sodic soil constraints and not primarily due to environmental variations. Hence, our study revealed that sodic soil constraints can intensify plant water stress. Statistical analysis between canopy temperature (9:30, 12:00, and 15:00 hrs) and crop biophysical parameters showed close negative correlations at both moderately sodic (R2 = 0.54 to 0.83) and highly sodic (R2 = 0.30 to 0.89) sites. A closer correlation was observed at 15:00 hrs for both sites. Thus, high-resolution UAV-thermal imaging has potential to detect water-stressed plants on sodic soil. Agglomerative hierarchical clustering was used as an unsupervised machine learning tool for ranking of physiological performance of wheat genotypes. Results suggest that UAV-thermal imaging and AHC techniques can discriminate cultivars tolerant to sodicity. The study improves our understanding of crop physiological behaviour and can assist farmers in selection of water stress tolerant genotypes to sustain food security in sodic soil under water-limited environments.

Levels of Illness Uncertainty Vary by Heart Failure Symptom Cluster Group

Background Heart failure patients experience multiple concurrent symptoms called symptom clusters. Understanding the role and characteristics of symptom clusters in heart failure (HF) patients is limited. Illness uncertainty may lead to difficulty interpreting symptoms and timely treatment. Before person-centered interventions can be developed, more information about variations in HF symptom clusters and factors associated with them are needed. Purpose The purpose of this study was to describe HF symptom clusters and determine if illness uncertainty levels differed by symptom cluster. Methods A total of 129 HF patients (mean age 76.4 (12.4) years, education 10.6 (3.2) years and time with HF 3.6 (5.9) years, 54% male, 93% Caucasian, 46% widowed/divorced, NYHA class III) completed the Heart Failure Somatic Perception Scale (HFSPS, 0-90, higher score indicates more symptoms) and Mishel's Uncertainty in Illness Scale (MUIS, 23-115, higher score indicates more uncertainty) during hospitalization. Symptom clusters were determined by hierarchical agglomerative clustering techniques in SPSS 25. After controlling for differences in age, gender and length of time with HF, ANCOVA (post hoc LSD) analyses explored uncertainty levels by symptom cluster group. Results Three clusters were found: 1. Shortness of breath cluster (SOB), n = 43 (fatigue, hard to breathe, SOB when dressing, orthopnea, DOE, PND, rest during day, wake urinate at night), 2. Edema cluster, n = 47 (swollen feet, tight shoes, tight clothes, weight gain), 3. Cardiac symptom cluster, n = 39 (palpitations, chest pain, upset stomach, decreased appetite, cough). After adjusting for age, gender and length of time with HF, there was a significant difference in uncertainty levels by symptom cluster group (p = .000). Post hoc analysis demonstrated that cardiac cluster members reported less illness uncertainty than edema cluster members (69.7 vs 75, respectively, p = .033). Conclusions Results suggest that significant differences exist in illness uncertainty levels based on the symptom experience of patients with HF. Care and management of HF symptoms should include a complete assessment of unique symptom cluster profiles.

Multivariate Analysis for Assessing Irrigation Water Quality: A Case Study of the Bahr Mouise Canal, Eastern Nile Delta

Water scarcity and suitable irrigation water management in arid regions represent tangible challenges for sustainable agriculture. The current study aimed to apply multivariate analysis and to develop a simplified water quality assessment using principal component analysis (PCA) and the agglomerative hierarchical clustering (AHC) technique to assess the water quality of the Bahr Mouise canal in El-Sharkia Governorate, Egypt. The proposed methods depended on the monitored water chemical composition (e.g., pH, water electrical conductivity (ECiw), Ca2+, Mg2+, Na+, K+, HCO3−, Cl−, and SO42−) during 2019. Based on the supervised classification of satellite images (Landsat 8 Operational Land Imager (OLI)), the distinguished land use/land cover types around the Bahr Mouise canal were agriculture, urban, and water bodies, while the dominating land use was agriculture. The water quality of the Bahr Mouise canal was classified into two classes based on the application of the irrigation water quality index (IWQI), while the water quality was classified into three classes using the PCA and AHC methods. Temporal variations in water quality were investigated, where the water qualities in winter, autumn, and spring (January, February, March, April, November, and December) were classified as class I (no restrictions) based on IWQI application, and the water salinity, sodicity, and/or alkalinity did not represent limiting factors for irrigation water quality. On the other hand, in the summer season (May, June, July, August, and October), the irrigation water was classified as class II (low restrictions); therefore, irrigation processes during summer may lead to an increase in the alkalinity hazard. The PCA classifications were compared with the IWQI results; the PCA classifications had similar assessment results during the year, except in September, while the water quality was assigned to class II using the PCA method and class I by applying the IWQI. Furthermore, the normalized difference vegetation index (NDVI) around the Bahr Mouise canal over eight months and climatic data assisted in explaining the fluctuations in water quality during 2019 as a result of changing the crop season and agriculture management. Assessments of water quality help to conserve soil, reduce degradation risk, and support decision makers in order to obtain sustainable agriculture, especially under water irrigation scarcity and the limited agricultural land in such an arid region.

A New Terrorism Categorization Based on Casualties and Consequences Using Hierarchical Clustering

This study deployed the agglomerative hierarchical clustering technique on the Global Terrorism Database to develop a new way of categorizing terrorism. Four clusters were extracted, resulting in a new categorization of terrorism based on casualties and consequences. The subcategories include low-casualty, medium-consequence terror groups; medium-casualty, high-consequence terror groups; high-casualty, low-consequence terror groups; and higher-casualty, low-consequence terror groups. The connectedness, separation, and compactness of the four clusters were evaluated and the results were reasonably accurate. This new categorization has useful application in counterterrorism regarding proportionate budgeting of weaponry and other resources.

Weightage of Mature Leaf in Eucalyptus Clones in Characterization and Variability Assessment for DUS Testing (Distinct, Uniform and Stable) using MuInvariant Analysis

The studies were undertaken to develop DUS testing guidelines of selected 25 Eucalyptus clones from commonly available 82 clones from Institute of Forest Genetics and Tree Breeding (IFGTB). The results revealed that all individual clones showed considerable variation in qualitative, pseudo-qualitative and quantitative characteristic traits (17 morpho-metric traits) of Mature leaves and were grouped into different categories. The traits like leaf shape, base shape, apex shape, margin, symmetry, petiole length, lamina area, length, breadth, perimeter, convex perimeter, curve length, curve width, roundness, aspect ratio, convex area and full ratio were identified as grouping traits. The weightage of the quantitative traits were proven by Multi-variant analysis. In Canonical Discriminant analysis, three significant discriminant functions, which had Eigen value &gt;1 and this 3 values accounted for a cumulative variation of 76.2%. In Principal Component Analysis, two significant principal components (PCs) extracted had Eigen value &gt;1 and out of 12 this two values accounted for a cumulative variation of 89.47%. In Eucalyptus clones the dendrogram drawn depicted three distinct clusters based on the similarity using Pearson correlation coefficient method. The dendrogram drawn depicted three distinct clusters based on the euclidean distances grouped them by dissimilarity using agglomerative hierarchical clustering technique (Ward’s method).

A generic hierarchical clustering approach for detecting bottlenecks in manufacturing

The advancements in machine learning (ML) techniques open new opportunities for analysing production system dynamics and augmenting the domain expert's decision-making. A common problem for domain experts on the shop floor is detecting throughput bottlenecks, as they constrain the system throughput. Detecting throughput bottlenecks is necessary to prioritise maintenance and improvement actions and obtain greater system throughput. The existing literature provides many ways to detect bottlenecks from machine data, using statistical-based approaches. These statistical-based approaches can be best applied in environments where the statistical descriptors of machine data (such as distribution of machine data, correlations and stationarity) are known beforehand. Computing statistical descriptors involves statistical assumptions. When the machine data doesn't comply with these assumptions, there is a risk of the results being disconnected from actual production system dynamics. An alternative approach to detecting throughput bottlenecks is to use ML- based techniques. These techniques, particularly unsupervised ML techniques, require no prior statistical information on machine data. This paper proposes a generic, unsupervised ML-based hierarchical clustering approach to detect throughput bottlenecks. The proposed approach is the outcome of systematic and careful selection of ML techniques. It begins by generating a time series of the chosen bottleneck detection metric and then clustering the time series using a dynamic time-wrapping measure and a complete-linkage agglomerative hierarchical clustering technique. The results are clusters of machines with similar production dynamic profiles, revealed from the historical data and enabling the detection of bottlenecks. The proposed approach is demonstrated in two real-world production systems. The approach integrates the concept of humans in-loop by using the domain expert's knowledge.

A robust hierarchical clustering for georeferenced data

The detection of spatially contiguous clusters is a relevant task in geostatistics since near located observations might have similar features than distant ones. Spatially compact groups can also improve clustering results interpretation according to the different detected subregions. In this paper, we propose a robust metric approach to neutralize the effect of possible outliers, i.e. an exponential transformation of a dissimilarity measure between each pair of locations based on non-parametric kernel estimator of the direct and cross variograms (Fouedjio, 2016) and on a different bandwidth identification, suitable for agglomerative hierarchical clustering techniques applied to data indexed by geographical coordinates. Simulation results are very promising showing very good performances of our proposed metric with respect to the baseline ones. Finally, the new clustering approach is applied to two real-word data sets, both giving locations and top soil heavy metal concentrations.

Variation in fruit quality traits and health promoting compounds of citrus fruits grown in semi-arid region

The present study was conducted to assess the variation in quality characteristics of nine commonly cultivated citrus genotypes grown in semi-arid region. Fruit weight varied from 130g (‘Minneola’) to 622g (‘Pummelo Red Flesh’) within different groups of citrus fruits. Juice content ranged from 31% (‘Pummelo Red Flesh’) to 54% (‘Mosambi’). Pummelo fruits were larger in size than fruits of other citrus groups. TSS varied from 8°B (‘Dancy’) to 11°B (‘Pummelo Red Flesh’), while ascorbic acid varied from 32 to 54 mg/100g. The maximum phenolic content was recorded in ‘Pummelo White Flesh’ fruits (44.66 mg GAE/100g) and the minimum was found in ‘Minneola’ (23.33 mg GAE/100g). The fruits of ‘Pummelo Red Flesh’ and ‘Pummelo White Flesh’ had highest flavonoids of 50.85 mg QE/100g and 48.10 mg QE/100g, respectively. The antioxidant activity was also high in the fruits of pummelo and grapefruit as compared to others. Principal Component Analysis revealed that TSS and acid content were the major quality attributes governing the citrus acceptability. Total flavonoid content and the in vitro FRAP antioxidant assay were the best to determine the functional quality of the fruits. Biplot analysis grouped the genotypes into two groups and the result was further verified by Agglomerative Hierarchical Clustering technique which too yielded similar results.

Comparative Study on Deep Convolution Neural Networks DCNN-Based Offline Arabic Handwriting Recognition

Recently, deep learning techniques demonstrated efficiency in building better performing machine learning models which are required in the field of offline Arabic handwriting recognition. Our ancient civilizations presented valuable handwritten manuscripts that need to be documented digitally. If we compared between Latin and the isolated Arabic character recognition, the latter is much more challenging due to the similarity between characters, and the variability of the writing styles. This paper proposes a multi-stage cascading system to serve the field of offline Arabic handwriting recognition. The approach starts with applying the Hierarchical Agglomerative Clustering (HAC) technique to split the database into partially inter-related clusters. The inter-relations between the constructed clusters support representing the database as a big search tree model and help to attain a reduced complexity in matching each test image with a cluster. Cluster members are then ranked based on our new proposed ranking algorithm. This ranking algorithm starts with computing Pyramid Histogram of Oriented Gradients (PHoG), and is followed by measuring divergence by Kullback-Leibler method. Eventually, the classification process is applied only to the highly ranked matching classes. A comparative study is made to assess the effect of six different deep Convolution Neural Networks (DCNNs) on the final recognition rates of the proposed system. Experiments are done using the IFN/ENIT Arabic database. The proposed clustering and ranking stages lead to using only 11% of the whole database in classifying test images. Accordingly, more reduced computation complexity and more enhanced classification results are achieved compared to recent existing systems.

Graph-based determination of structural controllability and observability for pressure and temperature dynamics during steam-assisted gravity drainage operation

Steam assisted gravity drainage (SAGD), which is used for the in-situ extraction and recovery of oil sands bitumen, is represented by a distributed parameter system (DPS). The problem of sensor placement and the control of steam chamber growth and oil production, respectively, require analysis of the observability and controllability of the system. In this type of system, parametric sensitivity is traditionally used in lieu of observability, and controllability has not been explored rigorously. In this work, we analyze the pressure and temperature fields of a SAGD model based on detailed reservoir simulations and present a data-driven technique to assess the structural controllability and observability of the system, with a view to determine optimal locations of actuators and sensors. An agglomerative hierarchical clustering technique is used to obtain a spanning tree of the clusters which is partitioned based on an objective function to arrive at a set of spatially contiguous clusters that display similar pressure/temperature dynamics. A Granger causality measure is used to create the linkage amongst the clusters to build a digraph model of the data. The driver nodes of the graph identify locations for actuation which provide full control over the graph, and the root strongly connected components indicate sensor locations which ensure structural observability over the entire graph. We demonstrate the method using data generated from SAGD simulations using the CMG-STARS simulator, identify the sensor and actuator locations required for complete structural observability and controllability of the system, and also provide a method of assessment of partial actuation and in-sensor ranges.

Unsupervised adaptation of PLDA models for broadcast diarization

We present a novel model adaptation approach to deal with data variability for speaker diarization in a broadcast environment. Expensive human annotated data can be used to mitigate the domain mismatch by means of supervised model adaptation approaches. By contrast, we propose an unsupervised adaptation method which does not need for in-domain labeled data but only the recording that we are diarizing. We rely on an inner adaptation block which combines Agglomerative Hierarchical Clustering (AHC) and Mean-Shift (MS) clustering techniques with a Fully Bayesian Probabilistic Linear Discriminant Analysis (PLDA) to produce pseudo-speaker labels suitable for model adaptation. We propose multiple adaptation approaches based on this basic block, including unsupervised and semi-supervised. Our proposed solutions, analyzed with the Multi-Genre Broadcast 2015 (MGB) dataset, reported significant improvements (16% relative improvement) with respect to the baseline, also outperforming a supervised adaptation proposal with low resources (9% relative improvement). Furthermore, our proposed unsupervised adaptation is totally compatible with a supervised one. The joint use of both adaptation techniques (supervised and unsupervised) shows a 13% relative improvement with respect to only considering the supervised adaptation.

Enhanced Slope One Algorithm using Hierarchical Clustering and Trust Based Collaborative Filtering for Ecommerce Applications

Collaborative filtering algorithm will be one among the assisting techniques delivering customized suggestions in the area of ecommerce. Nevertheless, conservative techniques concentrated in operating with client’s review and will not take into account of alteration of customer’s desires along with reliability of rankings associated. Huge quantity of increase in clients along with items resulted in certain critical complexities. Fresh Suggestion strategies will be required. Slope One algorithm might perform well with the motivation of minimized inadequacy of rankings, enhanced precision of suggestion. On the other hand increase in number of clients, resulted increased consumption duration. Establishment of solutions for complexities to extend adjacency space via utilization of clustering strategies will be carried out. Fundamental motivation of the paper relies with investigating feasible influence of utilizing trust measures in enhancing the quality of suggestions. This paper highlighted the significance of Trust in determining solutions for providing suggestions. Slope one algorithm incorporated with hierarchical agglomerative clustering technique performed superiorly while evaluated with trust metrics and solved the problem of huge amount of information associated with Trust aware information.

Efficient facial representations for age, gender and identity recognition in organizing photo albums using multi-output ConvNet.

This paper is focused on the automatic extraction of persons and their attributes (gender, year of born) from album of photos and videos. A two-stage approach is proposed in which, firstly, the convolutional neural network simultaneously predicts age/gender from all photos and additionally extracts facial representations suitable for face identification. Here the MobileNet is modified and is preliminarily trained to perform face recognition in order to additionally recognize age and gender. The age is estimated as the expected value of top predictions in the neural network. In the second stage of the proposed approach, extracted faces are grouped using hierarchical agglomerative clustering techniques. The birth year and gender of a person in each cluster are estimated using aggregation of predictions for individual photos. The proposed approach is implemented in an Android mobile application. It is experimentally demonstrated that the quality of facial clustering for the developed network is competitive with the state-of-the-art results achieved by deep neural networks, though implementation of the proposed approach is much computationally cheaper. Moreover, this approach is characterized by more accurate age/gender recognition when compared to the publicly available models.

Parallel Hierarchical Subspace Clustering of Categorical Data

Parallel clustering is an important research area of big data analysis. The conventional Hierarchical Agglomerative Clustering (HAC) techniques are inadequate to handle big-scale categorical datasets due to two drawbacks. First, HAC consumes excessive CPU time and memory resources; and second, it is non-trivial to decompose clustering tasks into independent sub-tasks executed in parallel. We solve these two problems by a MapReduce-based hierarchical subspace-clustering algorithm - called PAPU - using LSH-based data partitioning. PAPU is conducive to partitioning a large-scale dataset into multiple independent sub-datasets, into which similar data objects are mapped. Advocating parallel computing, PAPU obtains sub-clusters corresponding to respective attribute subspaces from independent chunks in the local clustering phase. To improve the accuracy of approximated clustering results, PAPU measures various scale clusters by applying the hierarchical clustering scheme to iteratively merge sub-clusters during the global clustering phase. We implement PAPU on a 24-node Hadoop computing platform. The experimental results reveal that hierarchical subspace-clustering coupled with the data-partitioning strategy achieves high clustering efficiency on both synthetic and real-world large-scale datasets. The experiments also demonstrate that PAPU delivers superior performance in terms of extensibility and scalability (e.g., a nearly linear speedup).

Rapid multi-objective optimization with multi-year future weather condition and decision-making support for building retrofit

A method of fast multi-objective optimization and decision-making support for building retrofit planning is developed, and lifecycle cost analysis method taking into account of future climate condition is used in evaluating the retrofit performance. In order to resolve the optimization problem in a fast manner with recourse to non-dominate sorting differential evolution algorithm, the simplified hourly dynamic simulation modeling tool SimBldPy is used as the simulator for objective function evaluation. Moreover, the generated non-dominated solutions are treated and rendered by a layered scheme using agglomerative hierarchical clustering technique to make it more intuitive and sense making during the decision-making process as well as to be better presented.The suggested optimization method is implemented to the retrofit planning of a campus building in UPenn with various energy conservation measures (ECM) and costs, and more than one thousand Pareto fronts are obtained and being analyzed according to the proposed decision-making framework. Twenty ECM combinations are eventually selected from all generated Pareto fronts. It is manifested that the developed decision-making support scheme shows robustness in dealing with retrofit optimization problem and is able to provide support for brainstorming and enumerating various possibilities during the decision-making process.

Visualization and Analysis in Bank Direct Marketing Prediction

Gaining the most benefits out of a certain data set is a difficult task because it requires an in-depth investigation into its different features and their corresponding values. This task is usually achieved by presenting data in a visual format to reveal hidden patterns. In this study, several visualization techniques are applied to a bank’s direct marketing data set. The data set obtained from the UCI machine learning repository website is imbalanced. Thus, some oversampling methods are used to enhance the accuracy of the prediction of a client’s subscription to a term deposit. Visualization efficiency is tested with the oversampling techniques’ influence on multiple classifier performance. Results show that the agglomerative hierarchical clustering technique outperforms other oversampling techniques and the Naive Bayes classifier gave the best prediction results.

Complexity Analysis Approach for Prefabricated Construction Products Using Uncertain Data Clustering

This paper proposes an uncertain data clustering approach to quantitatively analyze the complexity of prefabricated construction components through the integration of quality performance-based measures with associated engineering design information. The proposed model is constructed in three steps, which (1) measure prefabricated construction product complexity (hereafter referred to as product complexity) by introducing a Bayesian-based nonconforming quality performance indicator; (2) score each type of product complexity by developing a Hellinger distance-based distribution similarity measurement; and (3) cluster products into homogeneous complexity groups by using the agglomerative hierarchical clustering technique. An illustrative example is provided to demonstrate the proposed approach, and a case study of an industrial company in Edmonton, Canada, is conducted to validate the feasibility and applicability of the proposed model. This research inventively defines and investigates product complexity from the perspective of product quality performance with design information associated. The research outcomes provide simplified, interpretable, and informative insights for practitioners to better analyze and manage product complexity. In addition to this practical contribution, a novel hierarchical clustering technique is devised. This technique is capable of clustering uncertain data (i.e., beta distributions) with lower computational complexity and has the potential to be generalized to cluster all types of uncertain data.