Yazar "Aslan, Serpil" seçeneğine göre listele
Listeleniyor 1 - 20 / 22
Sayfa Başına Sonuç
Sıralama seçenekleri
Öğe A deep learning-based sentiment analysis approach (MF-CNN-BILSTM) and topic modeling of tweets related to the Ukraine-Russia conflict(Elsevier, 2023) Aslan, SerpilTwitter, one of the most significant social media platforms, can be used as data sources to research public opinion on various topics, including political conflicts. People worldwide have expressed their opinions about the war between Russia and Ukraine since it began. The motivation of this study is to use deep learning techniques to reveal qualitative and quantitatively narrow-scoped situational awareness and emotional tendencies during crisis periods. In order to achieve this, a sizable dataset of geotagged tweets was gathered with specific terms associated with the conflict between Ukraine and Russia. Then, deep learning-based text mining techniques like topic modeling and sentiment analysis were applied to investigate people's perspectives on conflict and emotional tendencies. The study will establish an impartial data source for the reports and articles of unbiased press members. This study used Valence Aware Dictionary and sEntiment Reasoner (VADER) to categorize the emotions related to the conflict between Russia and Ukraine in tweets. In addition, Latent Dirichlet Allocation (LDA) was used to extract various discussion topics. These techniques reveal the role of Twitter and compare and analyze the emotions and attitudes expressed on Twitter by different countries during the Ukraine-Russia conflict. In addition, a new deep learning-based sentiment classification MF-CNN-BiLSTMmodel that predicts and analyzes sentiments was proposed in this study. The proposed model stands for Multistage Feature Extraction using Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM). By combining various qualities and benefits of CNNs and BiLSTM, the suggested model makes it possible to identify both short-and long-term dependencies in ordinal data. The proposed model includes MF steps that strengthen feature extraction by identifying local dependencies. Experimental results demonstrate that it is possible to obtain more accurate sentiment classification results by the proposed method.& COPY; 2023 Elsevier B.V. All rights reserved.Öğe A Hybrid Artificial Intelligence Approach for Down Syndrome Risk Prediction in First Trimester Screening(Mdpi, 2025) Yalcin, Emre; Aslan, Serpil; Togacar, Mesut; Demir, Suleyman CansunBackground/Objectives: The aim of this study is to develop a hybrid artificial intelligence (AI) approach to improve the accuracy, efficiency, and reliability of Down Syndrome (DS) risk prediction during first trimester prenatal screening. The proposed method transforms one-dimensional (1D) patient data-including features such as nuchal translucency (NT), human chorionic gonadotropin (hCG), and pregnancy-associated plasma protein A (PAPP-A)-into two-dimensional (2D) Aztec barcode images, enabling advanced feature extraction using transformer-based deep learning models. Methods: The dataset consists of 958 anonymous patient records. Each record includes four first trimester screening markers, hCG, PAPP-A, and NT, expressed as multiples of the median. The DS risk outcome was categorized into three classes: high, medium, and low. Three transformer architectures-DeiT3, MaxViT, and Swin-are employed to extract high-level features from the generated barcodes. The extracted features are combined into a unified set, and dimensionality reduction is performed using two feature selection techniques: minimum Redundancy Maximum Relevance (mRMR) and RelieF. Intersecting features from both selectors are retained to form a compact and informative feature subset. The final features are classified using machine learning algorithms, including Bagged Trees and Naive Bayes. Results: The proposed approach achieved up to 100% classification accuracy using the Naive Bayes classifier with 1250 features selected by RelieF and 527 intersecting features from mRMR. By selecting a smaller but more informative subset of features, the system significantly reduced hardware and processing demands while maintaining strong predictive performance. Conclusions: The results suggest that the proposed hybrid AI method offers a promising and resource-efficient solution for DS risk assessment in first trimester screening. However, further comparative studies are recommended to validate its performance in broader clinical contexts.Öğe A Methodological Study on Fetal Health Classification Using Optimized LightGBM with SMOTE and Optuna-Based Hyperparameter Optimization(Turkiye Klinikleri, 2025) Yalcin, Emre; Aslan, Serpil; Tanyildiz, HayriyeObjective: Fetal health classification is of great clinical importance as it allows the early detection and management of fetal health problems during pregnancy.This study aims to enhance fetal health classification by integrating Light Gradient Boosting Machine (LightGBM), synthetic minority oversampling technique (SMOTE), and Optuna-based hyperparameter tuning. The goal is to improve classification accuracy, address class imbalance, and optimize model performance in predicting normal, suspicious, and pathological fetal health conditions. Material and Methods: The study utilized the University of California, Irvine Cardiotocography Data Set which contains 2.126 fetal cardiotocography (CTG) records classified into 3 categories. To handle class imbalance, SMOTE is applied. Various machine learning models [Stochastic Gradient Descent Classifier, Gradient Boosting, Decision Tree, Support Vector Machine, K-Nearest Neighbors, and Logistic Regression (LG)] were compared, with LightGBM selected due to its efficiency in structured medical data processing. Optuna is used for hyperparameter tuning. The dataset was split into 80% training and 20% testing, and model performance was assessed using accuracy, precision, recall, and F1-score. Results: The optimized LightGBM+SMOTE+Optuna model achieved 99% accuracy, significantly outperforming the other classifiers. The recall and F1-score for the suspicious and pathological classes improved, reducing the misclassification rates. The confusion matrix confirmed a substantial decrease in errors, demonstrating the model's robustness and reliability. Conclusion: The proposed model successfully enhances fetal health classification accuracy by addressing class imbalance and optimizing hyperparameters. The integration of LightGBM, SMOTE, and Optuna improves model generalization, making it a valuable tool for fetal health assessment. Future studies could explore deep learning approaches to further refine classification performance and support clinical decision-making.Öğe A new hybrid approach based on AOA, CNN and feature fusion that can automatically diagnose Parkinson's disease from sound signals: PDD-AOA-CNN(Springer London Ltd, 2024) Yildirim, Muhammed; Kiziloluk, Soner; Aslan, Serpil; Sert, EserParkinson's is one of the most rapidly increasing neurological diseases in the world, caused by the deficiency of dopamine-producing cells in the brain. Voice disorders are a significant finding in the early stage of Parkinson's disease (PD). Detection of this finding at an early stage of the disease allows early treatment of the disease. Therefore, in this study, using sound data, a hybrid model for detecting PD has been designed. In the developed method, first of all, the sound data were converted into spectrograms. Then, the feature maps of the obtained spectrogram images were extracted using 3 different CNN architectures. Feature maps with different features obtained by utilizing the accumulation of different architectures were combined. Then, these features were selected using the arithmetic optimization algorithm (AOA), one of the most recent metaheuristic optimization algorithms, and then classified by support vector machine (SVM) and K-nearest neighbors (KNN). One of the important novelties in the study is the reduction of the size of the acquired feature maps with AOA, a new and high-performance metaheuristic approach. The success of the proposed model in diagnosing Parkinson's disease reached up to 98.19%. In addition, feature maps of the sound data in the dataset were acquired by using the MFCC method to compare the performance of the proposed model. Eight different classifiers were used to categorize the acquired feature maps. The highest accuracy value obtained in this method was obtained in the Random Forest classifier with 93.98%.Öğe A New Hybrid Method for Classification of Rice Leaf Diseases: SVM+NCA+Resnet50(Niyazi BULUT, 2024) Bingöl, Harun; Aslan, SerpilRice is extremely important for individuals and countries, both in terms of nutritional value and financial value. It is necessary to protect such an important plant from diseases and increase the yield. However, early detection of diseases on plant leaves can prevent the spread of this disease and is also very important in terms of treating the plant. Artificial intelligence has become very popular in recent years thanks to its success in terms of disease classification. CNN architectures used in image classification perform very successful work. Within the scope of this study, it is recommended that the diseases on rice leaves be classified using artificial intelligence techniques, without mixing them with each other, with very high accuracy values, and without any problems caused by humans. With this proposed model, a support vector machine-based model is proposed that classifies five (5) of the most common rice diseases with a very high accuracy of %98.Öğe A novel attention-based deep learning model for improving sentiment classification after the case of the 2023 Kahramanmaras/Turkey earthquake on Twitter(Peerj Inc, 2025) Aslan, Serpil; Yildirim, MuhammedTwitter has emerged as one of the most widely used platforms for sharing information and updates. As users freely express their thoughts and emotions, a vast amount of data is generated, particularly in the aftermath of disasters, which can be collected quickly and directly from individuals. Traditionally, earthquake impact assessments have been conducted through field studies by non-governmental organizations (NGOs), a process that is often time-consuming and costly. Sentiment analysis (SA) on Twitter presents a valuable research area, enabling the extraction and interpretation of real-time public perceptions. In recent years, attention-based methods in deep learning networks have gained significant attention among researchers. This study proposes a novel sentiment classification model, MConv-BiLSTM-GAM, which leverages an attention mechanism to analyze public sentiment following the 7.8 and 7.5 Mw earthquakes that struck Kahramanmara & scedil;, Turkey. The model employs the FastText word embedding technique to convert tweets into vector representations. These vectorized inputs are then processed by a hybrid model integrating convolutional neural networks (CNNs) and recurrent neural networks (RNNs) with a global attention mechanism. This ensures careful consideration of semantic dependencies in sentiment classification. The proposed model operates in three stages: (i) MConv-Local Contextual Feature Extraction, (ii) bidirectional long short-term memory (BiLSTM)-sequence learning, and (iii) Global Attention Mechanism (GAM)-Attention Mechanism. Experimental results demonstrate that the model achieves an accuracy of 93.32%, surpassing traditional deep learning models in the literature by approximately 3%. This research aims to provide objective insights to policymakers and decision-makers, facilitating adequate support for individuals and communities affected by disasters. Moreover, analyzing public sentiment during earthquakes contributes to understanding societal responses and emotional trends in disaster scenarios.Öğe A novel feature fusion and mountain gazelle optimizer based framework for the recognition of jute pests in sustainable agriculture(Nature Portfolio, 2025) Kiziloluk, Soner; Karaduman, Mucahit; Aslan, Serpil; Yildirim, Muhammed; Khan, Muhammad Attique; Alhayan, Fatimah; Nam, YunyoungSustainable agriculture is an approach that involves adopting and developing agricultural practices to increase efficiency and preserve resources, both environmentally and economically. Jute is one of the primary sources of income grown in many countries. At this stage, increasing efficiency in jute production and protecting it from pests is essential. Detecting jute pests at an early stage will not only improve crop yield but also provide more income. In this paper, an artificial intelligence-based model was suggested to detect jute pests at an early stage. In this developed model, two different pre-trained models were used for feature extraction. To improve the performance of the developed model, the features obtained using the DarkNet-53 and DenseNet-201 models were combined. After this stage, the metaheuristic Mountain Gazelle Optimizer (MGO) was used, allowing the developed model to work faster and achieve more successful results. Feature selection was carried out using MGO; thus, more successful results were obtained with fewer, more compelling features. The proposed model was compared with six different models and five different classifiers accepted in the literature. In the developed model, 17 different jute pests were detected with 96.779% accuracy. The accuracy value achieved in the developed model is promising in successfully detecting jute pests.Öğe A novel TCNN-Bi-LSTM deep learning model for predicting sentiments of tweets about COVID-19 vaccines(Wiley, 2022) Aslan, SerpilMany researchers in various disciplines have focused on extracting meaningful information from social media platforms in recent years. Identification of behaviors and emotions from user posts is examined under the heading of sentiment analysis (SA) studies using the natural language processing (NLP) techniques. In this study, a novel TCNN-Bi-LSTM model using the two-stage convolutional neural network (TCNN) and bidirectional long short-term memory (Bi-LSTM) architectures was proposed. While TCNN layers enable the extraction of strong local features, the output of these layers feeds the Bi-LSTM model that remembers forward-looking information and capture long-term dependencies. In this study, first, preprocessing steps were applied to the raw dataset. Thus, strong features were extracted from the obtained quality dataset using the FastText word embedding technique that pre-trained with location-based and sub-word information features. The experimental results of the proposed method are promising compared to the baseline deep learning and machine learning models. Also, experimental results show that while the FastText data embedding technique achieves the best performance compared to other word embedding techniques in all deep learning classification models, it has not had the same outstanding success in machine learning models. This study aims to investigate the sentiments of tweets about the COVID-19 vaccines and comments on these tweets among Twitter users by using the power of Twitter data. A new dataset collected from Twitter was constructed to be used in experimental results. This study will facilitate detecting inappropriate, incomplete, and erroneous information about vaccination. The results of this study will enable society to broaden its perspective on the administered vaccines. It can also assist the government and healthcare agencies in planning and implementing the vaccination's promotion on time to achieve the herd immunity provided by the vaccination.Öğe Advancing multimodal emotion analysis: a hybrid deep learning approach with intermediate fusion and multi-task learning(Springer, 2026) Kaya, Fatih; Karaca, Yunus Emre; Aslan, Serpil; Yildirim, MuhammedEmotion analysis is a critical research domain focused on detecting the emotional states of individuals or communities across multiple data modalities, including text, images, and audio. While substantial progress has been made in unimodal (text-based) sentiment analysis, real-world scenarios often involve multimodal data, making integrated approaches essential for capturing contextual richness and improving predictive accuracy. This study introduces a hybrid deep learning model that combines text and visual features through an intermediate fusion mechanism and multi-task learning framework. Textual inputs are processed using RoBERTa and BiGRU layers, while visual inputs are analyzed through ViT and ResNet50 architectures enhanced by the Convolutional Block Attention Module (CBAM). The fused multimodal representations enable simultaneous and more robust emotion classification. Experimental results on the MVSA dataset demonstrate the superior performance of the proposed model, achieving 96.02% accuracy, 95.51% precision, 94.07% recall, and 94.73% F1-score, outperforming several state-of-the-art multimodal benchmarks. These findings underscore the model's methodological contributions and its strong potential for advancing the field of multimodal emotion analysis in both academic research and real-world applications.Öğe Automatic Classification of Particles in the Urine Sediment Test with the Developed Artificial Intelligence-Based Hybrid Model(Mdpi, 2023) Yildirim, Muhammed; Bingol, Harun; Cengil, Emine; Aslan, Serpil; Baykara, MuhammetUrine sediment examination is one of the main tests used in the diagnosis of many diseases. Thanks to this test, many diseases can be detected in advance. Examining the results of this test is an intensive and time-consuming process. Therefore, it is very important to automatically interpret the urine sediment test results using computer-aided systems. In this study, a data set consisting of eight classes was used. The data set used in the study consists of 8509 particle images obtained by examining the particles in the urine sediment. A hybrid model based on textural and Convolutional Neural Networks (CNN) was developed to classify the images in the related data set. The features obtained using textural-based methods and the features obtained from CNN-based architectures were combined after optimizing using the Minimum Redundancy Maximum Relevance (mRMR) method. In this way, we aimed to extract different features of the same image. This increased the performance of the proposed model. The CNN-based ResNet50 architecture and textural-based Local Binary Pattern (LBP) method were used for feature extraction. Finally, the optimized and combined feature map was classified at different machine learning classifiers. In order to compare the performance of the model proposed in the study, results were also obtained from different CNN architectures. A high accuracy value of 96.0% was obtained in the proposed model.Öğe Automatic Detection of Knee Osteoarthritis Disease with the Developed CNN, NCA and SVM Based Hybrid Model(Int Information & Engineering Technology Assoc, 2023) Aslan, SerpilKnee osteoarthritis (Knee-OA) is one of the most common musculoskeletal diseases caused by loss of cartilage and bone changes in the joint. Prediction of early Knee-OA based on early bone tissue analysis is challenging in medical image analysis. If the disease is detected in the later stages, it may cause serious problems, such as the need for knee replacement. Therefore, the detection of Knee-OA disease is essential. With the developing technology, computer-aided systems have been frequently used in the biomedical field in recent years. A deep learning-based hybrid model for the early diagnosis and treatment of Knee-OA disease was developed in this study. In the developed hybrid model, three different CNN architectures were used as the base, and feature extraction was made with these architectures. The features obtained in three different architectures are combined to bring together different features of the same image. After merging, the neighboring component analysis (NCA) size reduction method was used to remove unnecessary features. Since unnecessary features are eliminated from the feature map optimized with NCA, the proposed hybrid model will work faster and produce more successful results. Finally, the feature map optimized with NCA was classified with six different classifiers. The proposed model was also compared to eight different CNN architectures. In comparison to CNN architectures, the proposed hybrid model achieved the highest accuracy performance.Öğe Birbirine Benzeyen Üç Farklı Hastalığın Tespitinde Derin Öğrenme Modellerinin Performansı(Tokat Gaziosmanpaşa Üniversitesi, 2022) Karaca, Yunus Emre; Aslan, SerpilBir hastalığın doğru teşhis edilmesi ve doğru tedavi yöntemlerinin kullanması hastalıklı bireye kısa sürede şifayı sağlamada önemli iki kriterdir. Kısacası sorun bilinirse çözümü de kolaylaşacaktır. Çalışmamız da bu yine bu eksende olup gelişen tıp teknolojisini destekleyici mahiyettedir. Şöyle ki bir birine benzeyen üç hastalık tipi olan viral, bakteriyel ve COVID-19 pnömosine sahip hasta radyolojik görüntülerinin konvansiyonel sinir ağ(CNN) mimarileriyle hastalıkların tespit performanslarını karşılaştırdık. Bu karşılaştırmanın başarı oranın artması, doğru hastalık tanısı konulmasını da arttırmış olacaktır. Bu şekilde başarılı yöntemlerin ortaya çıkması hem teşhisi koyan hekimin işini kolaylaştırmasının yanı sıra tüm insanlık için en değerli kavram olan vakitten de tasarruf edilmiş olacak. 1281 COVID-19, 3270 Normal, 1656 viral-pnömoni ve 3001 bakteriyel-pnömonili toplamda 9208 göğüs röntgen görüntüsünün kullanıldığı çalışmamızda en başarılı performansı %88,05 ile Resnet50 mimarisi elde etmiştir.Öğe Content-Based Brain Magnetic Resonance Image Retrieval and Classification With the Proposed Deep Learning and Tissue-Based System(Ieee-Inst Electrical Electronics Engineers Inc, 2025) Dogan, Bedriye; Burak Mutlu, Hursit; Yildirim, Muhammed; Yalcin, Sercan; Aslan, Serpil; Sampathila, Niranjana; Rajendra Acharya, U.The exponential growth in the size of databases due to technological advancements has led to challenges in locating and accessing specific components of the data. While deep learning and other machine learning architectures have shown promise in retrieving data components, their efficacy is more pronounced when addressing disease cohorts. Contrarily, this effectiveness diminishes when accessing large datasets. This study focuses on the analysis of brain magnetic resonance imaging (MRI) images and, specifically, to differentiate between benign and malignant lesions associated with Alzheimer's disease, multiple sclerosis (MS), and intracranial regions, all of which are medically significant with distinct treatment modalities. A hybrid model was first devised to facilitate image retrieval by employing a pre-trained EfficientNet-b0 and local binary pattern (LBP) for feature extraction. These extracted features were then amalgamated to encompass diverse aspects of each image. To improve model performance, redundant features were pruned using the minimum redundancy maximum relevance (mRMR) technique. As a result, the proposed model demonstrated efficacy in analyzing a diverse dataset encompassing three distinct diseases and eight unique classes. Notably, existing machine architectures already published in the literature have struggled to achieve comparable success rates in discerning such closely related yet distinct disease groups. Our study underscores the challenge posed by increasing class diversity on the performance of deep learning architectures and obtained an accuracy of 98.9% in classifying three diseases and eight unique classes. As a result, the same model was used as the base in both the classification and CBIR processes for MRI detection, yielding competitive results when compared with the literature and other models.Öğe Detection of hateful twitter users with graph convolutional network model(Springer Heidelberg, 2023) Utku, Anil; Can, Umit; Aslan, SerpilToday, hate speech is widespread and persistent in various forms on social networking platforms, targeting different minority groups. These attacks can be carried out using various factors such as racial, religious, gender, and physical disability, etc. Considering the number of people and their interactions, social networks are the most important channels through which these discourses spread. The social network structure is considered a set of nodes and edges and is very suitable for the graph structure. The multidimensional structure of social networks carries social network data from Euclidean space to non-Euclidean space. In non-Euclidean space, the graph structure is used to represent data effectively. In this respect, solving the hate speech problem with graph-based methods in a complex dimensional space can produce more impressive results. In this study, a powerful method based on the Graph Convolutional Network (GCN) model, which is rarely used in this field, was proposed for the detection of hateful Twitter users in social networks. Well-known machine learning methods were used to measure the performance of this method. According to the results obtained, the proposed GCN model gave the most successful result.Öğe Detection of hateful twitter users with graph convolutional network model (vol 16, pg 329, 2023)(Springer Heidelberg, 2024) Utku, Anil; Can, Umit; Aslan, Serpil[Abstract Not Available]Öğe EPILEPTIC SEIZURE DETECTION FROM EEG SIGNALS WITH RECURRENT NEURAL NETWORKS BASED CLASSIFICATION MODEL(Niyazi BULUT, 2024) Aslan, Serpil; Bingöl, HarunEpileptic seizures are a neurological disorder that occurs as a result of sudden and uncontrolled electrical activities of the non-contagious brain. This condition may cause the person to lose normal activities temporarily. Epileptic seizures are a severe disease that affects approximately 60 million people in the world, usually manifested by symptoms such as loss of consciousness, muscle twitching, sudden sensory changes, or behavioural changes [1]. Genetics, brain injury, hormonal fluctuations, infections, or metabolic problems are some of the possible causes of epileptic seizures. Although the severity and duration of the seizure varies from person to person, it is usually very short and rarely reaches a point where it endangers human life. However, such seizures need to be recognized as soon as possible in order to improve the quality of life of individuals and reduce the frequency of seizures. Epileptic seizures are a manageable disease with early diagnosis and appropriate treatment. Recognizing epileptic seizures begins with understanding a person's symptoms and triggering factors. These symptoms may include loss of consciousness, muscle twitches, sudden sensory changes, and behavioural changes. The symptoms of seizures, past medical history, and neurological examinations are essential in the diagnosis process. From past to present, many methods have been developed for the early diagnosis and detection of epileptic seizures [2]. One of these is analyzing the brain's neural activities using electroencephalography (EEG), which helps experts make a diagnosis. Although EEG signals are used as a powerful tool in epileptic seizure recognition, distinguishing the signals within them is both costly and requires highly expert experience. Therefore, this study proposed an automatic classification model for pre-processed EEG signals using Dual-Tree Complex Wavelet Transform (DT-CWT) based on deep learning-based Recurrent Neural Networks (RNN) architecture to assist experts. Compared to classical machine learning methods, deep learning-based models require less manual feature engineering because they perform data automatically thanks to deep networks instead of manually selecting and transforming the data features. These advantages make the model more general and flexible. The proposed model aims to classify EEG signals and detect epileptic seizures effectively and quickly in the early stages.Öğe EVSEL ATIKLARIN DERİN ÖĞRENME TEKNİKLERİ İLE AYRIŞTIRILMASI(Ali KARCI, 2022) Karaca, Yunus Emre; Aslan, Serpil; Hark, CengizDerin öğrenme teknolojisinin hızlı gelişimi sayesinde günlük yaşantımızın hemen hemen her noktasında kullanılan akıllı sistemler geliştirilmektedir. Geliştirilen uygulamalar hayatımızı kolaylaştırdığı gibi doğaya da olumlu katkılar sağlamıştır. Geleneksel atık ayrıştırma yöntemleri, verimlilik ve doğruluk açısından yetersiz kalmaktadır. Ayrıca yüksek maliyetli olmasının yanında çevresel riskler bakımdan da sıkıntılar doğurabilir. Son yıllarda, yapay zekâ, makine öğrenmesi ve beraberinde getirdiği derin öğrenme teknikleri organik, evsel ve ambalaj atıkların ayrıştırılması gibi karmaşık problemlerin çözümünde popüler bir yöntem olmuştur. Bu çalışmada, hem insan ve canlı yaşamı hem de doğanın korunması açısından büyük öneme sahip olan evsel atıkların ayrıştırılması problemi ele alınmıştır. Yapay zekâ kümesinde yer alan makine öğrenmesinin bir alt kolu olan derin öğrenme ile evsel atıkların tespit edilip ayrıştırılması için popüler konvansiyonel sinir ağı (CNN) tabanlı ResNet-50, DenseNet-121, Inception-V3, VGG16 mimarileri kullanılarak sınıflandırma performansları karşılaştırılmıştır.Öğe Explainable tabular deep learning models for antenatal cesarean delivery prediction in multiparous women(Bmc, 2026) Yalcin, Emre; Tanyildiz, Hayriye; Aslan, Serpil; Demir, Suleyman Cansun; Sucu, Mete; Uzay, Fatma Islek; Bicer, AyseBackground/Objectives Globalincreases in cesarean section (C-section) rates, often exceeding medical necessity, highlight the need for accurate antenatal prediction to support evidence-based birth planning. Reliable prediction of delivery mode is essential for reducing maternal and neonatal morbidity, improving clinical decision-making, and optimizing resource allocation. This study analyzes a publicly available dataset of 460 multiparous women, including 18 obstetric and antenatal variables, published by Yimer and Mekonnen. Methods Deep learning architectures were systematically evaluated for predicting delivery mode in multiparous pregnancies. Classical Multilayer Perceptrons (MLPs) served as baseline models, while modern tabular deep learning methods were assessed as advanced alternatives. Preprocessing included multiple imputation, outlier removal, and class balancing via SMOTE. Feature selection was performed using a hybrid Boruta-clinical expert strategy. Hyperparameters were tuned through Random Search. To improve interpretability, an explainability pipeline integrating SHAP and LIME was incorporated. Results Optimized MLPs produced modest performance gains, but dedicated tabular models demonstrated clear superiority. TabNet achieved the highest performance, with an ROC-AUC of 0.79 and a PR-AUC of 0.74, attributed to its attention and masking mechanisms and robust handling of minority classes. TabPFN and CBAM-MLP yielded stable and balanced results, whereas FT-Transformer showed competitive yet comparatively moderate accuracy. Conclusions The findings demonstrate that modern tabular deep learning approaches, particularly TabNet, surpass baseline MLP architectures in terms of accuracy, explainability, and clinical applicability for predicting C-section in multiparous women. This study presents the first comprehensive and explainable comparison of tabular deep learning models tailored to multiparous pregnancies, combining hybrid Boruta-expert feature selection with SHAP and LIME interpretability. TabNet emerges as the most promising candidate for integration into clinical decision support systems, contributing substantially to Al-driven strategies for addressing rising global C-section rates.Öğe Hyperparameter optimization based machine learning approach for early diagnosis of fetal genetic disorders(Springer, 2025) Yalcin, Emre; Koc, Tarik Kaan; Aslan, Serpil; Demir, Suleyman Cansun; Aykut, Serdar; Sucu, MetePrenatal screening is the process of analyzing various clinical variables to estimate the risk of genetic disorders like Down Syndrome (DS), which is distinguished by intellectual disability, distinct facial features, and developmental delays. The accuracy of these risk assessments is heavily reliant on the suitability of the risk algorithm for the target population. This study proposes an enhanced machine learning (ML) approach for predicting Down Syndrome (DS) risk using first-trimester screening (FTS) data. The dataset includes clinical information from 959 women with singleton pregnancies at the & Ccedil;ukurova University Gynecology and Obstetrics Unit between 2020 and 2024. To address limitations in existing studies, GPT-4 was utilized to generate synthetic minority-class samples, and advanced feature engineering techniques were incorporated to enhance model robustness and interpretability. A predictive ML model was created, and Hyperparameter Tuning (HT) was applied to optimize it for performance. Eight classifiers were tested, and CatBoost performed the best, achieving 97.39% accuracy and a 2.62% false-positive rate, outperforming the second-best classifier (XGBoost) across all primary evaluation metrics. These improvements highlight the novelty of the framework, particularly its integration of GPT-4-based augmentation and engineered biochemical interaction features. The results demonstrate the model's potential for reliable DS risk prediction, offering a more efficient and less invasive alternative to traditional diagnostic procedures. By enhancing early risk detection, the method could reduce unnecessary referrals for invasive tests like amniocentesis, thereby minimizing patient anxiety and potential complications. Overall, the study contributes to the development of intelligent, data-driven solutions for prenatal care.Öğe Sentiment Analysis of Covid-19 Tweets by using LSTM Learning Model(Ali KARCI, 2021) Karaca, Yunus Emre; Aslan, SerpilSocial media plays an important role in our lives due to the conditions of the age we live. Nowadays, the most popular social media platform that prioritizes meaningful content sharing is Twitter. In Twitter, which produces big data on an unprecedented scale, users have the opportunity to share their own perspectives, feelings, and experiences, as well as examine the opinions of other individuals. The Coronavirus-2019 (Covid-19) disease, transmitted through close contact and small droplets spread by people coughing, sneezing, or speaking, has created social and economic wounds worldwide. As of July 7, 2021, more than 185 million people worldwide have been diagnosed with the New Coronavirus (Covid-19), and approximately 4 million people have died from this infectious disease. This work focuses on the analysis of the sentiments that Covid-19 leaves on people, using the tweets that people share about the Covid-19 pandemic on the Twitter platform. Analyzes are based on deep learning algorithms. Sentiment analysis can provide serious benefits. In this study, we used a Long-short Term Memory (LSTM) based network model. Also, we compared the proposed model other machine learning algorithms: Support Vector Machine (SVM), Naïve Bayes and Logistic Regression. Experimental results show that our proposed method can effectively perform sentiment analysis on the Twitter dataset.
