Browsing by Author "Asuroglu, Tunc"

Now showing 1 - 3 of 3

Citation - WoS: 3
Citation - Scopus: 4
A Comprehensive Evaluation of Oversampling Techniques for Enhancing Text Classification Performance
(Nature Portfolio, 2025) Taskiran, Salimkan Fatma; Turkoglu, Bahaeddin; Kaya, Ersin; Asuroglu, Tunc
Class imbalance is a common and critical challenge in text classification tasks, where the underrepresentation of certain classes often impairs the ability of classifiers to learn minority class patterns effectively. According to the "garbage in, garbage out" principle, even high-performing models may fail when trained on skewed distributions. To address this issue, this study investigates the impact of oversampling techniques, specifically the Synthetic Minority Over-sampling Technique (SMOTE) and thirty of its variants, on two benchmark text classification datasets: TREC and Emotions. Each dataset was vectorized using the MiniLMv2 transformer model to obtain semantically rich representations, and classification was performed using six machine learning algorithms. The balanced and imbalanced scenarios were compared in terms of F1-Score and Balanced Accuracy. This work constitutes, to the best of our knowledge, the first large-scale, systematic benchmarking of SMOTE-based oversampling methods in the context of transformer-embedded text classification. Furthermore, statistical significance of the observed performance differences was validated using the Friedman test. The results provide practical insights into the selection of oversampling techniques tailored to dataset characteristics and classifier sensitivity, supporting more robust and fair learning in imbalanced natural language processing tasks.
Citation - WoS: 6
Citation - Scopus: 6
A Novel Diversity Guided Galactic Swarm Optimization With Feedback Mechanism
(Ieee-Inst Electrical Electronics Engineers Inc, 2024) Uymaz, Oğuzhan; Türkoğlu, Bahaeddin; Kaya, Ersin; Asuroglu, Tunc
Galactic Swarm Optimization (GSO) is an optimization method inspired by the movements of stars and star clusters in the galaxy. This method aims to find the best solution in two phases using other known optimization methods. The first phase explores the search space, while the second phase tries to refine the best solution. In GSO, the population of the best individuals obtained from the first phase is used as the initial population for the second phase. This process is repeated until the stopping criterion is met. Although the knowledge obtained from the first phase is transferred to the second phase in GSO, there is no knowledge transfer from the second phase to the first phase. In this study, we propose a modification where the knowledge obtained in the second phase is transferred back to the first phase. Additionally, the Particle Swarm Optimization (PSO) method, used as the search method in the original study, has a fast convergence problem, which hinders an effective discovery process in the first phase of GSO. To address this, the proposed diversity-guided modification regulates population diversity and enhances exploration. To demonstrate the performance of the proposed method, twenty-six traditional benchmark functions and three engineering design problems were used. The proposed method was compared with the original GSO and six current optimization methods. The results of the experimental study were analyzed using statistical tests. The experimental results and analyses show that the proposed method performs successfully.
Citation - WoS: 5
Citation - Scopus: 5
Voice Analysis in Dogs With Deep Learning: Development of a Fully Automatic Voice Analysis System for Bioacoustics Studies
(Mdpi, 2024) Karaaslan, Mahmut; Turkoglu, Bahaeddin; Kaya, Ersin; Asuroglu, Tunc
Extracting behavioral information from animal sounds has long been a focus of research in bioacoustics, as sound-derived data are crucial for understanding animal behavior and environmental interactions. Traditional methods, which involve manual review of extensive recordings, pose significant challenges. This study proposes an automated system for detecting and classifying animal vocalizations, enhancing efficiency in behavior analysis. The system uses a preprocessing step to segment relevant sound regions from audio recordings, followed by feature extraction using Short-Time Fourier Transform (STFT), Mel-frequency cepstral coefficients (MFCCs), and linear-frequency cepstral coefficients (LFCCs). These features are input into convolutional neural network (CNN) classifiers to evaluate performance. Experimental results demonstrate the effectiveness of different CNN models and feature extraction methods, with AlexNet, DenseNet, EfficientNet, ResNet50, and ResNet152 being evaluated. The system achieves high accuracy in classifying vocal behaviors, such as barking and howling in dogs, providing a robust tool for behavioral analysis. The study highlights the importance of automated systems in bioacoustics research and suggests future improvements using deep learning-based methods for enhanced classification performance.