WoS İndeksli Yayınlar Koleksiyonu
Permanent URI for this collectionhttps://hdl.handle.net/20.500.12573/394
Browse
31 results
Search Results
Article High-Accuracy Identification of Durian Leaf Diseases: A Convolutional Neural Network Approach Validated with K-Fold Cross-Validation and Bayesian Optimization(Springer, 2025-11-18) Soylemez, Ismet; Nalici, Mehmet Eren; Unlu, RamazanTo address the economic losses caused by plant diseases in durian farming, this study presents an optimized deep learning model that diagnoses diseases from leaf images with high accuracy. The model's performance is maximized through Bayesian optimization and hyperparameter tuning, while its reliability is maximized through layered five-fold cross-validation. Training the convolutional neural network model on 2595 leaf images displaying six different states (five diseased and one healthy) resulted in an average test accuracy of 91.98%. This high, consistent success rate demonstrates the model's generalizability to different datasets without overfitting. While the 'Healthy' and 'Algal' classes were successfully detected with high F1-scores, there are difficulties distinguishing between the 'Blight' and 'Colletotrichum' classes due to visual similarities. This study establishes a new reference point for durian disease classification and makes a significant contribution to the development of reliable artificial intelligence-based diagnostic tools for precision agriculture.Conference Object Twist-Bend Instability of a Cantilever Beam Subjected to an End Load via Homotopy Perturbation Method(Amer Inst Physics, 2018) Yucesoy, Ahmet; Coskun, Safa Bozkurt; Atay, Mehmet Tarik; Cesoy, AhmetIn this article, twist-bend buckling analysis of a cantilever beam subjected to a concentrated end load is conducted using Homotopy Perturbation Method (HPM). Even in the linear stability analysis, obtaining an exact solution for some cases is not an easy task. However, by the use of HPM this difficulty can be overcome easily. This issue is presented with a case study and the results show that HPM can be used successfully in the analysis of twist-bend buckling of beams.Conference Object Citation - WoS: 2Citation - Scopus: 1The Numerical Solutions for Stiff Ordinary Differential Equations by Using Interpolated Variational Iteration Method With Comparison to Exact Solutions(Amer Inst Physics, 2018) Ciftci, Cihan; Cayci, Hatice Sinem Sas; Atay, Mehmet Tarik; Toker, Batuhan; Guncan, Berkay; Yildirim, Afsin TalhaRecently proposed Interpolated Variational Iteration Method (IVIM) is used to find numerical solutions of stiff ordinary differential equations for both linear and nonlinear problems. The examples are given to illustrate the accuracy and effectiveness of IVIM method and IVIM results are compared with exact results. In recent analytical approximate methods based studies related to stiff ordinary differential equations, problems were solved by Adomian Decomposition Method and VIM and Homotopy Perturbation Method, Homotopy Analysis Method etc. In this study comparisons with exact solutions reveal that the Interpolated Variational Iteration Method (IVIM) is easy to implement. In fact, this method is promising methods for various systems of linear and nonlinear stiff ordinary differential equations as an initial value problem. Furthermore, IVIM is giving very satisfactory solutions when compared to exact solutions for nonlinear cases depending on the stiffness ratio of the stiff system to be solved.Conference Object Text Classification Experiments on Contextual Graphs Built by N-Gram Series(Springer International Publishing AG, 2025) Sen, Tarik Uveys; Yakit, Mehmet Can; Gumus, Mehmet Semih; Abar, Orhan; Bakal, GokhanTraditional n-gram textual features, commonly employed in conventional machine learning models, offer lower performance rates on high-volume datasets compared to modern deep learning algorithms, which have been intensively studied for the past decade. The main reason for this performance disparity is that deep learning approaches handle textual data through the word vector space representation by catching the contextually hidden information in a better way. Nonetheless, the potential of the n-gram feature set to reflect the context is open to further investigation. In this sense, creating graphs using discriminative ngram series with high classification power has never been fully exploited by researchers. Hence, the main goal of this study is to contribute to the classification power by including the long-range neighborhood relationships for each word in the word embedding representations. To achieve this goal, we transformed the textual data by employing n-gram series into a graph structure and then trained a graph convolution network model. Consequently, we obtained contextually enriched word embeddings and observed F1-score performance improvements from 0.78 to 0.80 when we integrated those convolution-based word embeddings into an LSTM model. This research contributes to improving classification capabilities by leveraging graph structures derived from discriminative n-gram series.Conference Object TextNetTopics+: Enhancing Text Classification Through Classifier Diversity and Model Ensembling(Springer International Publishing AG, 2025) Voskergian, Daniel; Bakir-Gungor, Burcu; Yousef, MalikTextNetTopics 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 ensemblebased 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.Conference Object Citation - WoS: 3Citation - Scopus: 3Template Scoring Methods for Protein Torsion Angle Prediction(Springer-Verlag Berlin, 2015) Aydin, Zafer; Baker, David; Noble, William StaffordPrediction of backbone torsion angles provides important constraints about the 3D structure of a protein and is receiving a growing interest in the structure prediction community. In this paper, we introduce a three-stage machine learning classifier to predict the 7-state torsion angles of a protein. The first two stages employ dynamic Bayesian and neural networks to produce an ab-initio prediction of torsion angle states starting from sequence profiles. The third stage is a committee classifier, which combines the ab-initio prediction with a structural frequency profile derived from templates obtained by HHsearch. We develop several structural profile models and obtain significant improvements over the Laplacian scoring technique through: (1) scaling templates by integer powers of sequence identity score, (2) incorporating other alignment scores as multiplicative factors (3) adjusting or optimizing parameters of the profile models with respect to the similarity interval of the target. We also demonstrate that the torsion angle prediction accuracy improves at all levels of target-template similarity even when templates are distant from the target. The improvement is at significantly higher rates as template structures gradually get closer to target.Conference Object Temporal Logic-Based Intrusion Detection for Securing Connected Vehicles(Springer International Publishing AG, 2024) Bozdal, MehmetEnsuring the security and integrity of in-vehicle communication networks (IVCNs) is paramount. The increasing connectivity of vehicles exposes them to unprecedented security vulnerabilities, necessitating innovative methodologies to safeguard against cyberattacks and unauthorized access. This research presents a novel approach to enhance IVCN security through the deployment of a Signal Temporal Logic (STL)-based Intrusion Detection System (IDS). Considering the limited resources of Electronic Control Units (ECUs), this approach offers an adaptive and lightweight solution that addresses the unique challenges posed by the dynamic nature of vehicular networks. The proposed STL-based IDS effectively detects a broad spectrum of intrusions while maintaining acceptable overhead for resource-constrained ECUs, thanks to its distributed architecture. Comprehensive experimental evaluations demonstrate significant performance improvements in detecting Denial of Service (DoS) attacks, achieving the highest accuracy of 0.996 and recall of 1.000. The system also excels in detecting fuzzy attacks, with the highest accuracy of 0.996.Conference Object Target Attractor Tracking of Relative Phase in Bosonic Josephson Junction(Amer Inst Physics, 2016) Borisenok, SergeyThe relative phase of Bosonic Josephson junction in the Josephson regime of Bose-Hubbard model is tracked via the target attractor ('synergetic') feedback algorithm with the inter-well coupling parameter presented as a control function. The efficiency of our approach is demonstrated numerically for Gaussian and harmonic types of target phases.Conference Object Statistical Approach for Table Tennis Athletes' Success(Amer Inst Physics, 2018) Goren, Selcuk; Gulbahar, Ibrahim Tumay; Pinar, Muhammed SafakThis report summarizes the statistical modeling and analysis results associated with the athletes' success and athletes' features. Main purpose of this report is to find any relation between athletes' success and their features. As a tool of creating correlation regression is used with SPSS.Conference Object Citation - WoS: 17Citation - Scopus: 21Sol-Gel Applications for Ceramic Membrane Preparation(Amer Inst Physics, 2017) Erdem, I.Ceramic membranes possessing superior properties compared to polymeric membranes are more durable under severe working conditions and therefore their service life is longer. The ceramic membranes are composed of some layers. The support is the layer composed of coarser ceramic structure and responsible for mechanical durability under filtration pressure and it is prepared by consolidation of ceramic powders. The top layer is composed of a finer ceramic micro-structure mainly responsible for the separation of components present in the fluid to be filtered and sol-gel method is a versatile tool to prepare such a tailor-made ceramic filtration structure with finer pores. Depending on the type of filtration (e.g. micro-filtration, ultra-filtration, nano-filtration) aiming separation of components with different sizes, sols with different particulate sizes should be prepared and consolidated with varying precursors and preparation conditions. The coating of sol on the support layer and heat treatment application to have a stable ceramic micro-structure are also important steps determining the final properties of the top layer. Sol-gel method with various controllable parameters (e.g. precursor type, sol formation kinetics, heat treatment conditions) is a practical tool for the preparation of top layers of ceramic composite membranes with desired physicochemical properties.