Scopus İndeksli Yayınlar Koleksiyonu

Permanent URI for this collectionhttps://hdl.handle.net/20.500.12573/395

Browse

Filters

2018 - 201912020 - 20253

Settings

Has files: NoSubject: search.filters.subject.Features Selection

Search Results

Now showing 1 - 4 of 4
  • Article
    Citation - Scopus: 2
    Prediction of Colorectal Cancer Based on Taxonomic Levels of Microorganisms and Discovery of Taxonomic Biomarkers Using the Grouping-Scoring (G-S-M) Approach
    (Elsevier Ltd, 2025-03) Bakir-Güngör, Burcu; Temiz, Mustafa; Canakcimaksutoglu, Beyza; Yousef, Malik
    Colorectal cancer (CRC) is one of the most prevalent forms of cancer globally. The human gut microbiome plays an important role in the development of CRC and serves as a biomarker for early detection and treatment. This research effort focuses on the identification of potential taxonomic biomarkers of CRC using a grouping-based feature selection method. Additionally, this study investigates the effect of incorporating biological domain knowledge into the feature selection process while identifying CRC-associated microorganisms. Conventional feature selection techniques often fail to leverage existing biological knowledge during metagenomic data analysis. To address this gap, we propose taxonomy-based Grouping Scoring Modeling (G-S-M) method that integrates biological domain knowledge into feature grouping and selection. In this study, using metagenomic data related to CRC, classification is performed at three taxonomic levels (genus, family and order). The MetaPhlAn tool is employed to determine the relative abundance values of species in each sample. Comparative performance analyses involve six feature selection methods and four classification algorithms. When experimented on two CRC associated metagenomics datasets, the highest performance metric, yielding an AUC of 0.90, is observed at the genus taxonomic level. At this level, 7 out of top 10 groups (Parvimonas, Peptostreptococcus, Fusobacterium, Gemella, Streptococcus, Porphyromonas and Solobacterium) were commonly identified for both datasets. Moreover, the identified microorganisms at genus, family, and order levels are thoroughly discussed via refering to CRC-related metagenomic literature. This study not only contributes to our understanding of CRC development, but also highlights the applicability of taxonomy-based G-S-M method in tackling various diseases. © 2025 Elsevier B.V., All rights reserved.
  • Conference Object
    Metabolomics Data Analysis to Discover Chronic Granulomatous Disease-Associated Biomarkers Utilizing G-S-M Machine Learning Model via Grouping Metabolites According to Ion Type
    (Institute of Electrical and Electronics Engineers Inc., 2024-10-16) Ersöz, Nur Sebnem; Bakir-Güngör, Burcu; Yousef, Malik
    Chronic Granulomatous Disease (CGD) is a rare, inherited immunodeficiency disorder characterized by white blood cells unable to effectively kill certain bacteria and fungi. This defect results in the formation of clusters of immune cells called granulomas that form at sites of infection or inflammation. Therefore, identification of disease-related biomarkers is a critical step in advancing precision medicine and improving diagnostic accuracy. In this study, we applied a G-S-M machine learning approach to metabolomics data to uncover CGD-Associated biomarkers. We obtained a metabolomics dataset from Gene Expression Omnibus with GSE220260 accession number. Data includes 85 samples (16 healthy controls and 69 CGD samples) with comprehensive metabolic profiles obtained using liquid chromatography-mass spectrometry analysis. Dataset includes metabolite names with their ion type and formula. In order to identify CGD related metabolites and their ion types, G-S-M was used as a grouping function when performing machine learning oriented metabolomics data analysis. We have performed the G-S-M approach by grouping metabolites according to their ion type. In the training part of the G-S-M approach, metabolites annotated with selected ion types have been utilized to perform a two-class classification task which generates an important set of ion type output. We also compared the performance results of the G-S-M machine learning model with traditional feature selection methods; XGB, SKB, IG, FCBF, MRMR, CMIM with random forest classifier. 100 times Monte-Carlo Cross Validation was used in our experiments. It was observed that G-S-M, XGB, SKB and FCBF methods similarly provided the best performances. In this study, besides its performance, G-S-M method used groups based on ion types unlike TFS, and then identified relevant Chronic Granulomatous Disease-associated metabolites. © 2024 Elsevier B.V., All rights reserved.
  • Conference Object
    Citation - WoS: 22
    Citation - Scopus: 52
    Evaluation of Classification Algorithms, Linear Discriminant Analysis and a New Hybrid Feature Selection Methodology for the Diagnosis of Coronary Artery Disease
    (Institute of Electrical and Electronics Engineers Inc., 2018-12) Kolukisa, Burak; Hacilar, Hilal; Göy, Gökhan; Kus, Mustafa; Bakir-Güngör, Burcu; Aral, Atilla; Güngör, Vehbi Çağrı
    According to the World Health Organization (WHO), 31% of the world's total deaths in 2016 (17.9 million) was due to cardiovascular diseases (CVD). With the development of information technologies, it has become possible to predict whether people have heart diseases or not by checking certain physical and biochemical values at a lower cost. In this study, we have evalated a set of different classification algorithms, linear discriminant analysis and proposed a new hybrid feature selection methodology for the diagnosis of coronary heart diseases (CHD). Throughout this research effort, using three publicly available Heart Disease diagnosis datasets (UCI Machine Learning Repository), we have conducted comparative performance evaluations in terms of accuracy, sensitivity, specificity, F-measure, AUC and running time. © 2023 Elsevier B.V., All rights reserved.
  • Conference Object
    Enhancing Intrusion Detection in Electric Networks Using Physics-Informed Random Forest
    (Institute of Electrical and Electronics Engineers Inc., 2024-10-16) Bozdal, Mehmet; Savasci, Alper
    The increasing complexity of electric power networks has heightened their vulnerability to cyber-attacks, challenging traditional Intrusion Detection Systems (IDS) that rely on manually crafted rules. This paper introduces a novel approach that integrates physics-informed features and feature selection into a Random Forest (RF) model to enhance IDS performance. By deriving features such as complex power and impedance from fundamental electrical principles and applying SelectKBest for optimal feature selection, our method not only improves detection accuracy but also enhances efficiency by using fewer than half the features. Specifically, the feature-enriched RF model utilizing 55 features achieves an accuracy of 0.9667 and an F1-score of 0.9664, compared to 0.9576 and 0.9570 for the baseline RF model. This approach demonstrates the effectiveness of advanced feature engineering and selection techniques for improving the security and reliability of power network monitoring systems. © 2024 Elsevier B.V., All rights reserved.