Clustering-based Identification Research Articles

Clustering-Based Identification of BMI-Associated Metabolites with Mechanistic Insights from Network Analysis in Korean Men

Background: Epidemiological studies using metabolomics often encounter challenges due to metabolite profiles being influenced by multiple modifiable behavioral factors, including regular exercise, smoking, drinking, and weight control. This study aimed to identify modifiable behavioral factors reflected in metabolites by clustering subjects based on their metabolite profiles. Networks of metabolites were constructed to visualize their relationships and the differences between clustering groups. Methods: Sixty-four healthy men were included in this study. Information on regular exercise, smoking, and drinking was collected by questionnaires, and body mass index (BMI), an indicator of weight control, was calculated based on measured height and weight. Through targeted metabolomics, the concentrations of 149 metabolites were quantified. Subjects were clustered using the k-means method based on metabolite composition. Correlation-based networks were constructed for each cluster using Cytoscape software, followed by network analysis. Results: The subjects were divided into two clusters, with BMI identified as a distinguishing feature. Four lyso-phosphatidylcholines (PCs), six diacyl-PCs, and one acyl-alkyl-PC were positively associated with BMI. In the constructed network, acyl-alkyl-PCs exhibited the highest degrees, suggesting their central role in BMI-associated metabolic pathways. Conclusions: These findings suggest that metabolites can reflect behavioral factors, with BMI exerting a significant influence on metabolite profiles, particularly through its associations with phosphatidylcholines.

A clustering-based method for identifying and tracking squall lines

Abstract. The squall line is a type of convective system that is characterized by storm cells arranged in a line or band pattern and is usually associated with disastrous weather. The identification and tracking of squall lines thus play important roles in early warning systems for meteorological disasters. Here, a clustering-based identification and tracking algorithm for squall lines is presented based on weather radar data. A clustering analysis is designed to distinguish the strong echo area and estimate the feature value, including the reflectivity value, length, width, area, endpoints, central axes, and centroid. The linearly arranged clusters are merged to improve the identification of squall line development. The three-dimensional structure and movement tracking of the squall line are obtained using the centroid and velocity of the squall lines identified in each layer. The results demonstrate that the method can effectively identify and track one or more squall lines across the radar surveillance area. The results also show that the recognition accuracy rate for the single scan elevation of this method is 95.06 %, and the false-positive rate is 3.17 %. This method improves the accuracy of squall line identification in the development stage of squall lines and still works efficiently even when high interference contamination occurs.

Prediction of PM2.5 with a piecewise affine model considering spatial-temporal correlation

Over the past several decades, several air pollution prevention measures have been developed in response to the growing concern over air pollution. Using models to anticipate air pollution accurately aids in the timely prevention and management of air pollution. However, the spatial-temporal air quality aspects were not properly taken into account during the prior model construction. In this study, the distance correlation coefficient (DC) between measurements made in various monitoring stations is used to identify appropriate correlated monitoring stations. To derive spatial-temporal correlations for modeling, the causality relationship between measurements made in various monitoring stations is analyzed using Transfer Entropy (TE). This work explores the process of identifying a piecewise affine (PWA) model using a larger dataset and suggests a unique hierarchical clustering-based identification technique with model structure selection. This work improves the BIRCH (Balanced Iterative Reducing and Clustering using Hierarchies) by introducing Kullback-Leibler (KL) Divergence as the dissimilarity between clusters for handling clusters with arbitrary shapes. The number of clusters is automatically determined using a cluster validity metric. The task is formulated as a sparse optimization problem, and the model structure is selected using parameter estimations. Beijing air quality data is used to demonstrate the method, and the results show that the proposed strategy may produce acceptable forecast performance.

Extension of a clustering identification approach to multivariable piecewise affine systems: Application to an industrial dryer

The main objective of this paper is the extension of the clustering-based identification approach to Multi-Input Multi-Output (MIMO) PieceWise Affine systems (PWA). This approach is performed by three main steps which are data clustering, parameters matrices estimation and regions computing. Data clustering is the most important step because the performances depend on the results given by the used clustering algorithm. In the case of MIMO PWA systems, we should cluster matrices of parameters which are considered high dimensional data. However, most of the conventional clustering algorithms are not efficient since the similarity assessment which is based on the distances between objects is fruitless in high dimension space. Therefore, we propose an extension of the DBSCAN (Density Based Spatial Clustering of Applications with Noise) clustering technique to identify MIMO PWA systems. The simulation results presented in this paper illustrate the performance of the proposed method. An application to an industrial dryer of Di-Calcium Phosphate (DCP) is also presented in order to strengthen the simulation results.

On prediction of air pollutants with Takagi-Sugeno models based on a hierarchical clustering identification method

In recent years, air pollution has attracted considerable attention worldwide. As an effective air protection method, the accurate prediction of air pollutants can help provide an early warning against harmful air pollutants and plan air pollution prevention and control strategies. Due to its high prediction capability, a novel method using a Takagi-Sugeno fuzzy model is presented to predict air pollutants in this study. This work investigates fuzzy model identification based on a larger dataset, and a hierarchical clustering-based identification method with model structure selection is proposed. The novelty of this work is as follows: First, a novel hierarchical clustering method has been employed in air quality forecasting. This method improves the widely used BIRCH (Balanced Iterative Reducing and Clustering using Hierarchies) by introducing a refinement phase for handling clusters with arbitrary shapes. Second, a cluster validity measure automatically determines the number of clusters. Then, the estimation of the model order selection and the model parameters is formulated as a sparse optimization problem, which can be solved using global optimization. Finally, the proposed Takagi-Sugeno model can learn the local trend pattern and spatial-temporal dependencies of multivariate air quality-related time series data. The presented method is demonstrated with air quality measurements in Shanghai, and the results show that the proposed method can achieve acceptable prediction performance.

Whole-genome single nucleotide polymorphism analysis for typing the pandemic pathogen Fusarium graminearum sensu stricto.

Recent improvements in microbiology and molecular epidemiology were largely stimulated by whole- genome sequencing (WGS), which provides an unprecedented resolution in discriminating highly related genetic backgrounds. WGS is becoming the method of choice in epidemiology of fungal diseases, but its application is still in a pioneer stage, mainly due to the limited number of available genomes. Fungal pathogens often belong to complexes composed of numerous cryptic species. Detecting cryptic diversity is fundamental to understand the dynamics and the evolutionary relationships underlying disease outbreaks. In this study, we explore the value of whole-genome SNP analyses in identification of the pandemic pathogen Fusarium graminearum sensu stricto (F.g.). This species is responsible for cereal diseases and negatively impacts grain production worldwide. The fungus belongs to the monophyletic fungal complex referred to as F. graminearum species complex including at least sixteen cryptic species, a few among them may be involved in cereal diseases in certain agricultural areas. We analyzed WGS data from a collection of 99 F.g. strains and 33 strains representing all known cryptic species belonging to the FGSC complex. As a first step, we performed a phylogenomic analysis to reveal species-specific clustering. A RAxML maximum likelihood tree grouped all analyzed strains of F.g. into a single clade, supporting the clustering-based identification approach. Although, phylogenetic reconstructions are essential in detecting cryptic species, a phylogenomic tree does not fulfill the criteria for rapid and cost-effective approach for identification of fungi, due to the time-consuming nature of the analysis. As an alternative, analysis of WGS information by mapping sequence data from individual strains against reference genomes may provide useful markers for the rapid identification of fungi. We provide a robust framework for typing F.g. through the web-based PhaME workflow available at EDGE bioinformatics. The method was validated through multiple comparisons of assembly genomes to F.g. reference strain PH-1. We showed that the difference between intra- and interspecies variability was at least two times higher than intraspecific variation facilitating successful typing of F.g. This is the first study which employs WGS data for typing plant pathogenic fusaria.

Identification of piecewise affine model for batch processes based on constrained clustering technique

In this paper, a novel identification method of piecewise affine (PWA) model for batch processes based on constrained clustering technique is proposed. In traditional clustering-based identification approaches, data classification and region partition are performed individually so that inseparable problem usually occurs in the partition phase. The proposed method uses a constrained K-means clustering algorithm to simultaneously perform both data classification and region partition, which is accomplished by imposing the complete and non-overlapping partition constraints into the clustering optimization problem. We employ a greedy iterative approach combined with the golden section search to efficiently solve the constrained clustering problem. This method can greatly improve the accuracy of the identified PWA model. Finally, we demonstrate the effectiveness of the proposed identification method.

Mixture of Survival Analysis Models-Cluster-Weighted Weibull Distributions

Survival analysis is a widely used method to establish a connection between a time to event outcome and a set of variables. The goal of this work is to improve the accuracy of the widely applied parametric survival models. This work highlights that accurate and interpretable survival analysis models can be identified by clustering-based exploration of the operating regions of local survival models. The key idea is that when operating regions of local Weibull distributions are represented by Gaussian mixture models, the parameters of the mixture-of-Weibull model can be identified by a clustering algorithm. The proposed method is utilised in three case studies. The examples cover studying the dropout rate of university students, calculating the remaining useful life of lithium-ion batteries, and determining the chances of survival of prostate cancer patients. The results demonstrate the wide applicability of the method and the benefits of clustering-based identification of local Weibull models.

On identification of piecewise-affine models for systems with friction and its application to electro-mechanical throttles

Once dynamic process models are available, model-based analysis, design and testing methods for Hardware–in-the-Loop (HiL) simulation in context of developing motor control functions in automotive applications can be applied. For Hardware–in-the-Loop (HiL) simulation, a high-precision model is required. In this paper, a clustering-based identification method for piecewise-affine (PWA) models for systems with friction is presented that utilizes open-loop recorded data. The method is demonstrated with two industrial electro-mechanical throttles.

Automatic clustering-based identification of autoregressive fuzzy inference models for time series

We analyze the use of clustering methods for the automatic identification of fuzzy inference models for autoregressive prediction of time series. A methodology that combines fuzzy methods and residual variance estimation techniques is followed. A nonparametric residual variance estimator is used for a priori input and model selection. A simple scheme for initializing the widths of the input membership functions of fuzzy inference systems is proposed for the Improved Clustering for Function Approximation algorithm (ICFA), previously introduced for initializing RBF networks. This extension to the ICFA algorithm is shown to provide the most accurate predictions among a wide set of clustering algorithms. The method is applied to a diverse set of time series benchmarks. Its advantages in terms of accuracy and computational requirements are shown as compared to least-squares support vector machines (LS-SVM), the multilayer perceptron (MLP) and two variants of the extreme learning machine (ELM).

Clustering-based identification of clonally-related immunoglobulin gene sequence sets

BackgroundClonal expansion of B lymphocytes coupled with somatic mutation and antigen selection allow the mammalian humoral immune system to generate highly specific immunoglobulins (IG) or antibodies against invading bacteria, viruses and toxins. The availability of high-throughput DNA sequencing methods is providing new avenues for studying this clonal expansion and identifying the factors guiding the generation of antibodies. The identification of groups of rearranged immunoglobulin gene sequences descended from the same rearrangement (clonally-related sets) in very large sets of sequences is facilitated by the availability of immunoglobulin gene sequence alignment and partitioning software that can accurately predict component germline gene, but has required painstaking visual inspection and analysis of sequences. ResultsWe have developed and implemented an algorithm for identifying sets of clonally-related sequences in large human immunoglobulin heavy chain gene variable region sequence sets. The program processes sequences that have been partitioned using iHMMune-align, and uses pairwise comparisons of CDR3 sequences and similarity in IGHV and IGHJ germline gene assignments to construct a distance matrix. Agglomerative hierarchical clustering is then used to identify likely groups of clonally-related sequences. The program is available for download from http://www.cse.unsw.edu.au/~ihmmune/ClonalRelate/ClonalRelate.zip.ConclusionsThe method was evaluated on several benchmark datasets and provided a more accurate and considerably faster identification of clonally-related immunoglobulin gene sequences than visual inspection by domain experts.

Clustering-based identification for the prediction of splitting tensile strength of concrete

Splitting tensile strength (STS) of high-performance concrete (HPC) is one of the important mechanical properties for structural design. This property is related to compressive strength (CS), water/binder (W/B) ratio and concrete age. This paper presents a clustering-based fuzzy model for the prediction of STS based on the CS and (W/B) at a fixed age (28 days). The data driven fuzzy model consists of three main steps: fuzzy clustering, inference system, and prediction. The system can be analyzed directly by the model from measured data. The performance evaluations showed that the fuzzy model is more accurate than the other prediction models concerned.