Browsing by Author "Voskergian, Daniel"

Now showing 1 - 9 of 9

Correction: Engineering Novel Features for Diabetes Complication Prediction Using Synthetic Electronic Health Records
(Frontiers Media S.A., 2025) Voskergian, Daniel; Bakir-Gungor, Burcu; Yousef, Malik
Citation - Scopus: 1
Engineering Novel Features for Diabetes Complication Prediction Using Synthetic Electronic Health Records
(Frontiers Media S.A., 2025) Voskergian, Daniel; Bakir-Gungor, Burcu; Yousef, Malik
Diabetes significantly affects millions of people worldwide, leading to substantial morbidity, disability, and mortality rates. Predicting diabetes-related complications from health records is crucial for early prevention and for the development of effective treatment plans. In order to predict four different complications of diabetes mellitus, i.e., retinopathy, chronic kidney disease, ischemic heart disease, and amputations, this study introduces a novel feature engineering approach. While developing the classification models, we utilize XGBoost feature selection method and various supervised machine learning algorithms, including Random Forest, XGBoost, LogitBoost, AdaBoost, and Decision Tree. These models were trained on synthetic electronic health records (EHR) generated by dual-adversarial autoencoders. These EHRs represent nearly 1 million synthetic patients derived from an authentic cohort of 979,308 individuals with diabetes. The variables considered in the models were the age range accompanied by chronic diseases that occur during patient visits starting from the onset of diabetes. Throughout the experiments, XGBoost and Random Forest demonstrated the best overall prediction performance. The final models, which are tailored to each complication and trained using our feature engineering approach, achieved an accuracy between 69% and 77% and an AUC between 77% and 84% using cross-validation, while the partitioned validation approach yielded an accuracy between 59% and 78% and an AUC between 66% and 85%. These findings imply that the performance of our method surpass the performance of the traditional Bag-of-Features approach, highlighting the effectiveness of our approach in enhancing model accuracy and robustness.
ETNT: Enhanced Textnettopics With Filtered LDA Topics and Sequential Forward / Backward Topic Scoring Approaches
(Science & information Sai Organization Ltd, 2024) Voskergian, Daniel; Jayousi, Rashid; Bakir-Gungor, Burcu
TextNetTopics is a novel text classification-based topic modelling approach that focuses on topic selection rather than individual word selection to train a machine learning algorithm. However, one key limitation of TextNetTopics is its scoring component, which evaluates each topic in isolation and ranks them accordingly, ignoring the potential relationships between topics. In addition, the chosen topics may contain redundant or irrelevant features, potentially increasing the feature set size and introducing noise that can degrade the overall model performance. To address these limitations and improve the classification performance, this study introduces an enhancement to TextNetTopics. eTNT integrates two novel scoring approaches: Sequential Forward Topic Scoring (SFTS) and Sequential Backward Topic Scoring (SBTS), which consider topic interactions by assessing sets of topics simultaneously. Moreover, it incorporates a filtering component that aims to enhance topics' quality and discriminative power by removing non-informative features from each topic using Random Forest feature importance values. These integrations aim to streamline the topic selection process and enhance classifier efficiency for text classification. The results obtained from the WOS-5736, LitCovid, and MultiLabel datasets provide valuable insights into the superior effectiveness of eTNT compared to its counterpart, TextNetTopics.
Citation - Scopus: 1
eTNT: Enhanced Textnettopics With Filtered LDA Topics and Sequential Forward / Backward Topic Scoring Approaches
(Science and Information Organization, 2024) Voskergian, Daniel; Jayousi, Rashid; Bakir-Güngör, Burcu
TextNetTopics is a novel text classification-based topic modelling approach that focuses on topic selection rather than individual word selection to train a machine learning algorithm. However, one key limitation of TextNetTopics is its scoring component, which evaluates each topic in isolation and ranks them accordingly, ignoring the potential relationships between topics. In addition, the chosen topics may contain redundant or irrelevant features, potentially increasing the feature set size and introducing noise that can degrade the overall model performance. To address these limitations and improve the classification performance, this study introduces an enhancement to TextNetTopics. eTNT integrates two novel scoring approaches: Sequential Forward Topic Scoring (SFTS) and Sequential Backward Topic Scoring (SBTS), which consider topic interactions by assessing sets of topics simultaneously. Moreover, it incorporates a filtering component that aims to enhance topics' quality and discriminative power by removing non-informative features from each topic using Random Forest feature importance values. These integrations aim to streamline the topic selection process and enhance classifier efficiency for text classification. The results obtained from the WOS-5736, LitCovid, and MultiLabel datasets provide valuable insights into the superior effectiveness of eTNT compared to its counterpart, TextNetTopics. © 2024 Elsevier B.V., All rights reserved.
Citation - Scopus: 1
Semant - Feature Group Selection Utilizing Fasttext-Based Semantic Word Grouping, Scoring, and Modeling Approach for Text Classification
(Springer International Publishing AG, 2024) Voskergian, Daniel; Bakir-Gungor, Burcu; Yousef, Malik
Text classification presents a challenge due to its high-dimensional feature space. As such, devising an effective feature selection scheme is essential. In this study, we present SEMANT, a novel hybrid filter-wrapper feature selection method that utilizes filter-based Chi-Square and the wrapper-based G-S-M approach. SEMANT incorporates fastText neural word embedding similarities to promote greater semantic inclusion in the selection of features for text classification tasks. The performance of the proposed method was investigated on the WOS-5736 and LitCovid datasets and compared with TextNetTopics, a topic modeling-based topic selection algorithm for text classification. Experimental results confirm that the proposed approach outperforms its alternative.
Citation - WoS: 10
Citation - Scopus: 15
Textnettopics Pro, a Topic Model-Based Text Classification for Short Text by Integration of Semantic and Document-Topic Distribution Information
(Frontiers Media S.A., 2023) Voskergian, Daniel; Bakir-Gungor, Burcu; Yousef, Malik
With the exponential growth in the daily publication of scientific articles, automatic classification and categorization can assist in assigning articles to a predefined category. Article titles are concise descriptions of the articles' content with valuable information that can be useful in document classification and categorization. However, shortness, data sparseness, limited word occurrences, and the inadequate contextual information of scientific document titles hinder the direct application of conventional text mining and machine learning algorithms on these short texts, making their classification a challenging task. This study firstly explores the performance of our earlier study, TextNetTopics on the short text. Secondly, here we propose an advanced version called TextNetTopics Pro, which is a novel short-text classification framework that utilizes a promising combination of lexical features organized in topics of words and topic distribution extracted by a topic model to alleviate the data-sparseness problem when classifying short texts. We evaluate our proposed approach using nine state-of-the-art short-text topic models on two publicly available datasets of scientific article titles as short-text documents. The first dataset is related to the Biomedical field, and the other one is related to Computer Science publications. Additionally, we comparatively evaluate the predictive performance of the models generated with and without using the abstracts. Finally, we demonstrate the robustness and effectiveness of the proposed approach in handling the imbalanced data, particularly in the classification of Drug-Induced Liver Injury articles as part of the CAMDA challenge. Taking advantage of the semantic information detected by topic models proved to be a reliable way to improve the overall performance of ML classifiers.
TextNetTopics+: Enhancing Text Classification Through Classifier Diversity and Model Ensembling
(Springer Science and Business Media Deutschland GmbH, 2025) Voskergian, Daniel; Bakir-Güngör, Burcu; Yousef, Malik
TextNetTopics is an innovative text classification framework that integrates topic modeling with feature selection to improve model accuracy and interpretability. Unlike traditional methods that rely on individual words, TextNetTopics selects cohesive topics extracted via Latent Dirichlet Allocation as features for document representation, effectively reducing dimensionality while preserving the semantic structure of the text. This study evaluates the performance of TextNetTopics utilizing multiple machine learning algorithms in the M (Modeling) component, including Random Forest, Support Vector Machine, Gradient Boosting, eXtreme Gradient Boosting, and Logistic Regression. To further enhance classification performance, we introduce TextNetTopics+, an ensemble-based extension that leverages both hard voting and soft voting mechanisms to combine the strengths of multiple classifiers. Comprehensive experiments on the LitCovid and WOS datasets demonstrate that ensemble learning in TextNetTopics + significantly outperforms individual classifiers in TextNetTopics, confirming its effectiveness in improving model robustness and generalization. © 2025 Elsevier B.V., All rights reserved.
Citation - WoS: 1
Citation - Scopus: 1
Textnettopics-SFTS-SBTS Textnettopics Scoring Approaches Based Sequential Forward and Backward
(Springer International Publishing AG, 2024) Voskergian, Daniel; Bakir-Gungor, Burcu; Yousef, Malik
TextNetTopics is a text classification-based topic modeling approach that performs topic selection rather than word selection to train a machine learning algorithm. However, one main limitation of TextNetTopics is that its scoring component (the S component) assesses each topic independently and ranks them accordingly, neglecting the potential relationship between topics. In order to address this limitation and improve the classification performance, this study introduces an enhancement to TextNetTopics. TextNetTopics-SFTS-SBTS integrates two novel scoring approaches: Sequential Forward Topic Scoring (SFTS) and Sequential Backward Topic Scoring (SBTS), which consider topic interactions by assessing sets of topics simultaneously. This integration aims to streamline the topic selection process and enhance classifier efficiency for text classification. The results obtained across three datasets offer valuable insights into the context-dependent effectiveness of the new scoring mechanisms across diverse datasets and varying numbers of topics involved in the analysis.
Citation - Scopus: 1
TextNetTopics_TIS: Enhancing Textnettopics With Random Forest-Based Topic Importance Scoring
(Institute of Electrical and Electronics Engineers Inc., 2024) Voskergian, Daniel; Bakir-Güngör, Burcu; Yousef, Malik
TextNetTopics is an innovative Latent Dirichlet Allocation-based topic selection method for training text classification models. One main limitation is its computationally intensive scoring mechanism, especially when applied to many topics. This scoring mechanism involves training a machine learning model (i.e., Random Forest) on each topic using the Monte-Carlo Cross-Validation approach and assigning a score value based on a specific performance metric (e.g., accuracy or F1-score). Moreover, the measured score does not account for the interactions between all features residing in all topics. This paper presents a new topic-scoring mechanism called Topic Importance Scoring. This computationally efficient approach trains a Random Forest model on all topics simultaneously and leverages the extracted feature importance values to give each topic a score reflecting its classification potential. The experiments on three diverse datasets confirm that the proposed method's performance is superior to the Topic Performance Scoring, which was used in the original TextNetTopics method. © 2024 Elsevier B.V., All rights reserved.