SEARCH · Search PubMed

Results for “network analysis”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 559 records · Page 31Linked to original sources

PubMed2025

VEGFA sex-specific signature is associated to long COVID symptom persistence.

BACKGROUND: Long COVID involves persistent symptoms after COVID-19 recovery, affecting multiple organ systems for months or years. Risk factors include female sex, prior chronic conditions, severe SARS-CoV-2 infection, reinfections, and lack of vaccination. As a major public health concern, ongoing research continues to investigate its causes, mechanisms, and long-term effects. METHODS: Proteomic expression analysis of 171 individuals, in two time points, with confirmed SARS-CoV-2 infection, including 133 long COVID patients from the deeply characterized COVICAT cohort, assessed 1395 protein biomarkers using Olink® technology. Statistical analyses with linear mixed models examined protein expression changes, long COVID status, and sex-specific differences. Functional analysis included gene set enrichment analysis and protein-protein interaction networks. RESULTS: Findings revealed VEGFA overexpression in long COVID patients (effect size 0.322, SE = 0.098, p = 0.0013), along with sex-specific expression patterns and the influence of sex-hormonal status in females, with significant overexpression of circulating VEGFA levels specifically in postmenopausal women (Mann-Whitney U test p value = 8.55 × 10-3). Network analysis identified 109 nodes and 274 edges, with VEGFA ranking highest in centrality. Dysregulated chemokine signaling, complement activation, and viral reactivation were also confirmed, consistent with prior studies. CONCLUSIONS: Using high-throughput proteomic profiling in a population-based cohort, we observed that vascular dysfunction, particularly involving VEGFA, is a key feature of long COVID, especially in milder cases, with significant overexpression of VEGFA in postmenopausal women. Sex-specific proteomic patterns suggest distinct recovery mechanisms, highlighting the need to consider sex, vascular health, and disease severity in the pathogenesis and management of long COVID.

Humans↗

PubMed2026

LINC01871-Mediated Sensitivity to Cyclin-Dependent Kinase 4/6 Inhibitors in Human Breast Cancer.

Breast cancer remains the most frequently diagnosed malignancy in women, and resistance to cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors limits long-term treatment efficacy. This study aimed to identify long non-coding RNAs (lncRNAs) associated with predicted sensitivity to CDK4/6 inhibitors and to investigate their biological functions in breast cancer. Transcriptomic data from The Cancer Genome Atlas (TCGA) and drug sensitivity data from the Genomics of Drug Sensitivity in Cancer 2 (GDSC2) database were integrated, and drug sensitivity was predicted using the oncoPredict algorithm. Candidate lncRNAs were identified through differential expression analysis, weighted gene co-expression network analysis, prognostic analysis, and machine learning. The biological functions of LINC01871 were subsequently evaluated using in vitro and in vivo experiments. Sixty-two lncRNAs associated with predicted sensitivity to ribociclib and palbociclib were identified, and six core lncRNAs were selected. LINC01871 showed the highest discriminatory performance for predicted drug sensitivity. Overexpression of LINC01871 was associated with increased sensitivity of breast cancer cells to ribociclib and palbociclib, inhibition of cell proliferation, promotion of apoptosis, and suppression of nuclear factor kappa B (NF-κB) signaling. Single-cell transcriptomic analysis demonstrated high LINC01871 expression in T cells and natural killer (NK) cells, while transcriptome-based immune infiltration analyses showed that high LINC01871 expression was associated with increased immune infiltration. These findings identify LINC01871 as a candidate biomarker of sensitivity to CDK4/6 inhibitors and demonstrate its tumor-suppressive effects in breast cancer. Further clinical and mechanistic studies are required to validate its predictive value and therapeutic relevance.

Humans↗

PubMed2026

Distinct periarticular muscle transcriptomes: inflammation in rheumatoid arthritis versus metabolic dysregulation in osteoarthritis.

OBJECTIVES: Periarticular skeletal muscle abnormalities are recognised in rheumatoid arthritis (RA) and osteoarthritis (OA), but their divergent molecular pathologies are poorly defined. This study aimed to elucidate and directly compare the transcriptomic profiles of periarticular muscle in patients with RA and OA. METHODS: We performed bulk RNA sequencing of periarticular skeletal muscle samples collected during total joint arthroplasty from RA (n=6) and OA (n=4) patients. Differential gene expression analysis, weighted gene co-expression network analysis (WGCNA), pathway enrichment, and gene set variation analyses were conducted to identify disease-specific molecular features and their clinical associations. RESULTS: The two conditions showed fundamentally distinct profiles. RA muscle exhibited a pronounced inflammatory signature, characterised by upregulation of cytokine-responsive genes including FOS, EGR1, and CXCL2, and enrichment of tumour necrosis factor-α and interleukin-6 (IL-6)/JAK-STAT3 signalling. In contrast, OA muscle was characterised by metabolic dysregulation, with upregulation of genes linked to adipogenesis (PCK1, SFRP4) and significant enrichment of epithelial-to-mesenchymal transition (EMT) signalling. These divergent profiles were further supported by WGCNA, which identified distinct modules reflecting heightened innate immune and complement activation in RA, and disrupted metabolic processes in OA. Notably, in RA, the IL-2-STAT5 signalling pathway was unique among those tested in showing a strong positive correlation with DAS28-ESR (r=0.94, p=0.019). CONCLUSIONS: This study reveals distinct molecular pathologies in the periarticular muscle of RA and OA. RA muscle shows an intense inflammatory profile potentially linked to cachexia, whereas OA muscle displays features of metabolic disease and pro-fibrotic remodelling.

Humans↗

PubMed1994

Three faces of integrative coordination: a model of interorganizational relations in community-based health and human services.

OBJECTIVE: This study develops a theoretically justified, network-based model of integrative coordination in community-based health and human services, and it uses this model to measure and compare coordination in six elder service systems. DATA SOURCES AND STUDY SETTING: We collected data between 1989 and 1991 in six Alabama counties, including two major MSAs, two small MSAs, and two rural areas. STUDY DESIGN AND DATA COLLECTION/EXTRACTION METHODS: Our measurement of coordination is based on patterns of interorganizational relationships connecting the agencies constituting a community-based health and human services system. Within each site, we interviewed representatives from these agencies, asking them to indicate client referral, generalized support, and agenda-setting relationships they had developed with each of the other agencies in the system. Using network analysis procedures we then identified the network associated with each of these organizational functions (i.e., service delivery, administration, and planning) in each site, and we assessed levels of coordination in each network. PRINCIPAL FINDINGS: Our measure of integrative coordination is consistent with other indicators of coordination we derive from our data, suggesting its validity. In addition, levels of integrative coordination across sites for each organizational function are generally comparable. Comparisons across sites show integrative coordination to be consistently highest for service delivery networks and lowest for planning networks. CONCLUSIONS: Previous attempts to assess interorganizational coordination without regard to organizational function are subject to misinterpretation. The differing interorganizational dynamics involved in service delivery, administration, and planning appear to generate different patterns of interorganizational relationships, and different levels of coordination.

Aged↗

PubMed2025

Comprehensive analysis of diagnostic biomarkers related to histone acetylation in acute myocardial infarction.

BACKGROUND: Acute myocardial infarction (AMI) has become a serious disease that endangers human health, with high morbidity and mortality. Numerous studies have reported histone acetylation can result in the occurrence of cardiovascular diseases. This article aims to explore the potential biomarkers of histone acetylation regulatory genes (ARGs) in AMI patients. METHODS: Five AMI datasets were downloaded from the Gene Expression Omnibus (GEO) database. Next, ARG-related genes were gathered by gene set variation analysis (GSVA) and Spearman's correlation analysis. Subsequently, weighted gene co-expression network analysis (WGCNA) was performed to identify the module genes related to histone acetylation regulation. In the GSE60993 and GSE48060 datasets, the common differentially expressed genes (DEGs) between AMI and control samples were screened. Importantly, the intersecting genes were obtained by overlapping ARGs-related genes, common DEGs, and module genes. Then, the biomarkers in AMI were determined by machine learning, receiver operating characteristic (ROC) curves, and quantitative PCR (qPCR). In addition, immune analysis, drug prediction, molecular docking, and the lncRNA-miRNA-mRNA regulatory network targeting the biomarkers were analyzed, respectively. RESULTS: Here, a total of 18 intersecting genes were identified by overlapping 7,349 ARGs-related genes, 5,565 module genes, and 25 common DEGs. Further, five biomarkers (AQP9, HLA-DQA1, MCEMP1, NKG7, and S100A12) were obtained, and a nomogram was constructed and verified based on these biomarkers. Notably, the biomarkers were significantly associated with CD8 T cells and neutrophils. In addition, the drugs related to biomarkers were predicted, and ATOGEPANT with the molecular target (S100A12) had a high binding affinity (docking score = -10 kcal/mol). CONCLUSION: AQP9, HLA-DQA1, MCEMP1, NKG7, and S100A12 were identified as biomarkers related to ARGs in AMI, which provides a new perspective to study the relationship between ARGs and AMI.

Humans↗

PubMedDate not supplied

A novel neural network-based survival analysis model.

A feedforward neural network architecture aimed at survival probability estimation is presented which generalizes the standard, usually linear, models described in literature. The network builds an approximation to the survival probability of a system at a given time, conditional on the system features. The resulting model is described in a hierarchical Bayesian framework. Experiments with synthetic and real world data compare the performance of this model with the commonly used standard ones.

Bayes Theorem↗

PubMed1993

Application of neural networks in the QSAR analysis of percent effect biological data: comparison with adaptive least squares and nonlinear regression analysis.

Artificial neural networks (ANN) can be used for the direct QSAR analysis of percent effect biological data, thus avoiding the bias introduced by arbitrarily chosen classes and the loss of information due to prior classification. For two data sets the ANN results are compared with those obtained by adaptive least squares and nonlinear regression analyses. In comparison with the other methods the neural network shows higher predictive power and does not require an explicit equation relating the observed effect to physicochemical descriptors.

Animals↗

PubMed2006

H-CORE: enabling genome-scale Bayesian analysis of biological systems without prior knowledge.

The Bayesian network is a popular tool for describing relationships between data entities by representing probabilistic (in)dependencies with a directed acyclic graph (DAG) structure. Relationships have been inferred between biological entities using the Bayesian network model with high-throughput data from biological systems in diverse fields. However, the scalability of those approaches is seriously restricted because of the huge search space for finding an optimal DAG structure in the process of Bayesian network learning. For this reason, most previous approaches limit the number of target entities or use additional knowledge to restrict the search space. In this paper, we use the hierarchical clustering and order restriction (H-CORE) method for the learning of large Bayesian networks by clustering entities and restricting edge directions between those clusters, with the aim of overcoming the scalability problem and thus making it possible to perform genome-scale Bayesian network analysis without additional biological knowledge. We use simulations to show that H-CORE is much faster than the widely used sparse candidate method, whilst being of comparable quality. We have also applied H-CORE to retrieving gene-to-gene relationships in a biological system (The 'Rosetta compendium'). By evaluating learned information through literature mining, we demonstrate that H-CORE enables the genome-scale Bayesian analysis of biological systems without any prior knowledge.

Algorithms↗

PubMed2026

Beyond one-to-one mappings: Modelling distributed lesion-symptom relationships with multilayer networks.

Lesion-symptom mapping is widely used to identify causal relationships between brain structures and behaviour, and has played a central role in neuropsychologically informed network models of cognition. However, even recent approaches remain constrained by a one-to-one mapping framework, which oversimplifies the complex relationships between network-level damage and cognitive deficits. In addition, the non-orthogonality of cortical and white matter damage makes it difficult to disentangle their distinct contributions. Here, we used graph-based multilayer network analysis to address these limitations and evaluate clinical relevance. Using neuroanatomical and longitudinal neuropsychological data from 252 patients who underwent awake neurosurgery for low-grade glioma, we constructed interactive, three-layer networks for each hemisphere. Layer 1 comprised neuropsychological tasks (NT), layer 2 structural disconnections (SD), and layer 3 cortical damage (CD). Nodes represented tasks, white matter tracts, and cortical parcels, respectively, whereas within-layer edges captured correlations in performance or co-occurring damage patterns. Multilayer community detection identified domain- and hemisphere-specific brain-behaviour motifs linking executive, language, and spatial functions to distinct combinations of cortical and white matter disruption, a pattern confirmed by two spatial embedding approaches. Centrality analyses revealed a continuum of mapping relationships, ranging from one-to-one to one-to-many associations, indicating that tasks such as verbal fluency are better explained by multiple disconnection mechanisms. Additional analyses uncovered many-to-one and many-to-many relationships and highlighted tracts and cortical regions with domain-general relevance. Together, these findings support a neurobiologically grounded, network-oriented account of how structural brain damage gives rise to cognitive deficits, with implications for clinical care.

Humans↗

PubMed2004

Selecting compounds for focused screening using linear discriminant analysis and artificial neural networks.

Linear discriminant analysis and a committee of neural networks have been applied to recognise compounds that act at biological targets belonging to a specific gene family, protein kinases. The MDDR database was used to provide compounds targeted against this family and sets of randomly selected molecules. BCUT parameters were employed as input descriptors that encode structural properties and information relevant to ligand-receptor interactions. The technique was applied to purchasing compounds from external suppliers. These compounds achieved hit rates on a par with those achieved using known actives for related targets when tested for the ability to inhibit kinases at a single concentration. This approach is intended as one of a series of filters in the selection of screening candidates, compound purchases and the application of synthetic priorities to combinatorial libraries.

Databases, Factual↗

PubMed2026

KLF5 promotes proliferation, migration, and autophagy-/EMT‑associated molecular changes in lens epithelial cells via transcriptional activation of THBS1 in traumatic cataract.

PURPOSE: Traumatic cataract is a common blinding eye disease after ocular trauma, and its pathogenesis is closely related to lens epithelial cell dysfunction, while the definite molecular regulatory mechanism between upstream transcription factor and downstream target gene remains poorly clarified. This study aimed to clarify the role and molecular mechanism of the krüppel-like factor 5 (KLF5)/ thrombosponin 1 (THBS1) axis in regulating proliferation, migration, epithelial-mesenchymal transition and autophagy of lens epithelial cells in traumatic cataract, and to explore its potential clinical therapeutic value. METHODS: The GSE295383 dataset in the gene expression omnibus (GEO) database was downloaded, and the differentially expressed genes (DEGs) were screened by linear models for microarray data (limma) package of R language. Combined with Weighted gene co-expression network analysis (WGCNA), the gene co-expression network was constructed and the key modules were screened. Gene ontology (GO), kyoto encyclopedia of genes and genomes (KEGG) and gene set enrichment analysis (GSEA) combined with human transcription factor target (hTFtarget) and JASPAR databases were used to predict the upstream transcription factors of THBS1. Subsequently, SRA01/04 cells were induced with transforming growth factor-beta 2 (TGF-β2) to construct a cataract cell model. RESULTS: THBS1 and KLF5 were highly expressed in LECs exposed to TGF-β2. KLF5 could activate THBS1 transcription by binding to THBS1 promoter - 174 to -165 sites. Knockdown of THBS1 inhibited TGF-β2-induced viability, proliferation, migration, and altered the expression of epithelial-mesenchymal transition (EMT)- and autophagy-related markers in LECs. Knockdown of KLF5 downregulated THBS1 expression and produced a similar inhibitory effect, while overexpression of THBS1 reversed the effect of KLF5 knockdown. CONCLUSIONS: This study demonstrated that KLF5 promoted the proliferation, migration, and EMT‑associated molecular changes of LECs in traumatic cataract through transcriptional activation of THBS1, and regulated the expression of autophagy‑related markers in LECs, suggesting that KLF5/THBS1 axis might be a potential target for the treatment of traumatic cataract.

Cataract↗

PubMed2026

Symptom Networks and Core Symptoms in Patients with Solid Tumors Undergoing Chemotherapy: A Systematic Review.

OBJECTIVES: To summarize symptom network characteristics in patients with solid tumors undergoing chemotherapy and synthesize evidence on core symptoms, bridge symptoms, and temporal associations. METHODS: We systematically searched eight databases through October 2025 to identify studies that applied symptom network analysis to adults with solid tumors receiving chemotherapy. Eligible studies assessed symptoms using cross-sectional, longitudinal, or interventional designs. Two reviewers independently screened articles and extracted data on study characteristics, symptom assessment, and network outcomes. Methodological quality was assessed using the National Institutes of Health Study Quality Assessment Tool. RESULTS: Twenty-seven studies involving 13,452 participants were included, yielding 79 symptom networks. Fatigue was the most frequently identified core symptom (10/20, 50%), whereas sadness, lack of appetite, and nausea each occurred in 10% of studies, with variation across cancer types, treatment phases, and latent classes. Bridge symptoms included disturbed sleep, lack of appetite, and dry mouth (2/7, 28.6%). Studies evaluating temporal associations found that symptoms such as sadness, dyspnea, somnolence, and dry mouth predicted subsequent changes in appetite, distress, nausea, and other outcomes. Strength metrics showed acceptable stability (correlation stability coefficients: 0.28-0.83). CONCLUSIONS: Fatigue was frequently identified as a central symptom across studies, largely reflecting evidence from breast cancer studies. Core symptoms varied across cancer types, treatment phases, and latent classes, suggesting heterogeneity. IMPLICATIONS FOR NURSING PRACTICE: These findings highlight the importance of considering relationships among symptoms in clinical care. Focusing on key symptoms such as fatigue, while tailoring management strategies to cancer-specific symptom patterns, may support more effective symptom management.

Humans↗

PubMed1997

Application of neural networks to the analysis of pyrolysis mass spectra.

Pyrolysis mass spectrometry is a data rich analysis technique now becoming widely applied in microbiology. Data analysis is a key step in the exploitation of the technique and the application of neural network analysis to pyrolysis mass spectrometric data offers new opportunities for classification, identification and inter-strain comparison of microorganisms in biotechnology and clinical microbiology. The use of a supervised neural network for the identification of members of a streptomycete species-group is described.

Mass Spectrometry↗

PubMedDate not supplied

Event Analysis of Systemic Teamwork (EAST): a novel integration of ergonomics methods to analyse C4i activity.

C4i is defined as the management infrastructure needed for the execution of a common goal supported by multiple agents in multiple locations and technology. In order to extract data from complex and diverse C4i scenarios a descriptive methodology called Event Analysis for Systemic Teamwork (EAST) has been developed. With over 90 existing ergonomics methodologies already available, the approach taken was to integrate a hierarchical task analysis, a coordination demand analysis, a communications usage diagram, a social network analysis, and the critical decision method. The outputs of these methods provide two summary representations in the form of an enhanced operation sequence diagram and a propositional network. These offer multiple overlapping perspectives on key descriptive constructs including who the agents are in a scenario, when tasks occur, where agents are located, how agents collaborate and communicate, what information is used, and what knowledge is shared. The application of these methods to live data drawn from the UK rail industry demonstrates how alternative scenarios can be compared on key metrics, how multiple perspectives on the same data can be taken, and what further detailed insights can be extracted. The ultimate aim of EAST is, by applying it across a number of scenarios in different civil and military domains, to provide data to develop generic models of C4i activity and to improve the design of systems aimed at enhancing this management infrastructure.

Communication↗

PubMed2026

Quantitative Proteomic Analysis of APP/PS1 Transgenic Mice.

BACKGROUND: Alzheimer's disease (AD) is a prevalent neurodegenerative disorder affecting the central nervous system (CNS), with its etiology still shrouded in uncertainty. The interplay of extracellular amyloid-β (Aβ) deposition, intracellular neurofibrillary tangles (NFTs) composed of tau protein, cholinergic neuronal impairment, and other pathogenic factors is implicated in the progression of AD. OBJECTIVE: The current study endeavors to delineate the proteomic landscape alterations in the hippocampus of an AD murine model, utilizing proteomic analysis to identify key physiological and pathological shifts induced by the disease. This endeavor aims to shed light on the underlying pathogenic mechanisms, which could facilitate early diagnosis and pave the way for novel therapeutic interventions for AD. METHODS: To dissect the proteomic perturbations induced by Aβ and Presenilin-1 (PS1) in the AD pathogenesis, we undertook a label-free quantitative (LFQ) proteomic analysis focusing on the hippocampal proteome of the APP/PS1 transgenic mouse model. Employing a multi-faceted approach that included differential protein functional enrichment, cluster analysis, and protein-protein interaction (PPI) network analysis, we conducted a comprehensive comparative proteomic study between APP/PS1 transgenic mice and their wild-type C57BL/6 counterparts. RESULTS: Mass spectrometry identified a total of 4817 proteins in the samples, with 2762 proteins being quantifiable. Comparative analysis revealed 396 proteins with differential expression between the APP/PS1 and control groups. Notably, 35 proteins exhibited consistent temporal regulation trends in the hippocampus, with concomitant alterations in biological pathways and PPI networks. CONCLUSIONS: This study presents a comparative proteomic profile of transgenic (APP/PS1) and wild-type mice, highlighting the proteomic divergences. Furthermore, it charts the trajectory of proteomic changes in the AD mouse model across the developmental stages from 2 to 12 months, providing insights into the physiological and pathological implications of the disease-associated genetic mutations.

Animals↗

PubMed1998

Community-based trauma systems in the United States: an examination of structural development.

OBJECTIVE: To examine the organizational, political, and community characteristics that facilitate or impede community progress in developing a coordinative network of health services for trauma delivery. STUDY SETTING/DESIGN: A comparative case study design was used to examine trauma network development in 6 U.S. cities with a population of 1,000,000 or more. Five key coordinative activities were selected for study. Each study site varied in the set of activities that had been implemented. DATA SOURCES: Information on the structure and composition of local trauma coordinating councils; interviews with a common set of informants in each site using a semi-structured interview protocol. STUDY METHODS: The literature on interorganizational community structures and local policy development was drawn upon to create a conceptual framework for assessing the development of a coordinative service network. Analytical techniques included network analysis to understand the linkages across organizations in overseeing trauma network operations, assessment of leadership structures to identify central actors and organizations, and pattern matching techniques of case study analysis to identify factors that affected trauma network development. PRINCIPAL FINDINGS: Leaders capitalized on local events and were instrumental in keeping network development on the top of the political agenda. Successful leaders spent substantial time and energy documenting problems, assessing the needs and understanding of stakeholders, educating stakeholders and politicians, and creating trust and shared understanding of values. CONCLUSIONS: Prior research has documented the importance of central actors and organizations in developing coordinative networks. The unique contribution of our research is its insights on how central actors and organizations are more likely to motivate collaboration in situations where they lack control over the allocation of payments across involved organizations. Our research suggests that under these circumstances central players should focus their time and energy educating stakeholders and developing a shared understanding rather than using their centrality to impose a particular coordinative structure. To date, U.S. trauma networks have served as models for other industrialized countries, and thus, lessons learned in the U.S. about implementing interorganizational networks of trauma care can assist other countries achieve more effective coordination and avoid mistakes that impede progress.

Community Health Planning↗

PubMed2026

Malaria Risk among Internally Mobile Individuals and Heterogeneous Mobility Patterns in Two Hypoendemic Communities: Implications for Malaria Elimination in the Peruvian Amazon.

BACKGROUND: Human mobility is increasingly recognized as a key factor influencing malaria transmission dynamics, particularly in low-transmission settings approaching elimination. This study aimed to assess mobility patterns and their association with malaria risk in two hypoendemic communities in the Peruvian Amazon. METHOD: A longitudinal study was conducted in the communities of Libertad and Urcomiraño (Mazán River basin). Monthly population screenings were combined with weekly active and passive case detection. A total of 678 individuals were enrolled. Mobility patterns were assessed through structured questionnaires, and social network analysis was used to characterize travel connections. Log-binomial regression analysis was applied to identify risk factors associated with malaria infection. RESULT: Internally, mobile individuals in Libertad showed a higher malaria incidence (>32.47 cases per 1,000 person-months) than those in Urcomiraño (<10.15 cases per 1,000 person-months). Travel networks were mainly connected to Mazan district and Iquitos city, followed by local streams such as Armas and Arahuana. Mobility was primarily driven by family, administrative and occupational activities. Male sex (PR = 2.15, 95% CI: 1.37 - 3.37) and age ≥15 years (PR = 1.98, 95% CI: 1.24 - 3.19) were significantly associated with malaria infection (p-value < 0.05). CONCLUSION: Internally mobile populations represent a key high-risk group sustaining malaria transmission in hypoendemic settings. Targeted interventions focusing on mobile individuals should be integrated into malaria elimination strategies in the Peruvian Amazon and similar endemic regions.

Malaria elimination↗

PubMed1997

A pattern classification procedure integrating the multivariate statistical analysis with neural networks.

A new procedure integrating multivariate statistical analysis with artificial neural networks (ANN) for complex pattern classification is proposed. Firstly, a specially designed statistical analysis algorithm called correlative component analysis (CCA) was used to identify the classification characteristics (CC) from original high-dimensional pattern information. These CC were then used as input data to the ANN for pattern classification. The proposed new procedure not only effectively decreased the dimensionality of original patterns, but also took advantage of the self-learning power of the ANN. Further, a typical example of classifying natural spearmint essence was employed to verify the effectiveness of the new pattern classification method. The study showed that this novel integrated procedure provides better results than those obtained using individual methods separately.

Algorithms↗