212 research outputs found
Epilepsy diagnosis using artificial neural network learned by PSO
No full textElectroencephalogram (EEG) is used routinely for diagnosis of diseases occurring in the brain. It is a very useful clinical tool in the classification of epileptic seizures and the diagnosis of epilepsy. In this study, epilepsy diagnosis has been investigated using EEG records. For this purpose, an artificial neural network (ANN), widely used and known as an active classification technique, is applied. The particle swarm optimization (PSO) method, which does not need gradient calculation, derivative information, or any solution of differential equations, is preferred as the training algorithm for the ANN. A PSO-based neural network (PSONN) model is diversified according to PSO versions, and 7 PSO-based neural network models are described. Among these models, PSONN3 and PSONN4 are determined to be appropriate models for epilepsy diagnosis due to having better classification accuracy. The training methods-based PSO versions are compared with the backpropagation algorithm, which is a traditional method. In addition, different numbers of neurons, iterations/generations, and swarm sizes have been considered and tried. Results obtained from the models are evaluated, interpreted, and compared with the results of earlier works done with the same dataset in the literature
Azerbaijan and Azerbaijanis in the short story book of Cihan Aktaş titled
No full textDüşünce, inceleme kitaplarıyla olduğu kadar öyküleriyle de tanınan Cihan Aktaş, 1992 – 2009 yılları arasında dokuz öykü kitabı yayımlamıştır. Özellikle Azize’nin Son Günü (1997) adlı öykü kitabında olmak üzere, öykülerinde Azerbaycan’dan ve Azeri Türklerinden bahseden yazar, bu metinlerde gerek Azeri Türklerinden ilginç portreleri edebiyatımıza kazandırmakta, gerekse Azerbaycan kültürüyle, folkloruyla, sosyal ve siyasi hayatıyla, ekonomik şartlarıyla ilgili önemli bilgiler aktarmaktadır. Ayrıca başta Baku olmak üzere Azerbaycan coğrafyasına ve çevre bölgelere ilişkin izlenimlerini öyküleştirmektedir. Bu makalede, Cihan Aktaş’ın Azize’nin Son Günü adlı öykü kitabına yansıyan Azerbaycan ve Azeri Türkü imgesi/algısı incelenmeye çalışılacaktır.Cihan Aktaş who is known by his books of thought and review as well as short stories published nine short story books between 1992 and 2009. The author, who refers to Azerbaijan and Azerbaijanis, especially in his short story book titled the Last Day of Azize (1997), either brought interesting portraits of Azerbaijani Turks in these texts or conveyed important information related to the culture, folklore, social and political life and economical conditions of Azerbaijan. He also narrated his impressions related to Azerbaijan geography and environmental area, especially to Baku. In this article, we will attempt to understand image/perception of Azerbaijan and Azerbaijanis reflecting on the book of Cihan Aktaş titled “the Last Day of Azize”
Cihan Aktaş’ın "Azize’nin Son Günü" adlı öykü kitabında Azerbaycan ve Azeriler
No full textDüşünce, inceleme kitaplarıyla olduğu kadar öyküleriyle de tanınan Cihan Aktaş, 1992 – 2009 yılları arasında dokuz öykü kitabı yayımlamıştır. Özellikle Azize’nin Son Günü (1997) adlı öykü kitabında olmak üzere, öykülerinde Azerbaycan’dan ve Azeri Türklerinden bahseden yazar, bu metinlerde gerek Azeri Türklerinden ilginç portreleri edebiyatımıza kazandırmakta, gerekse Azerbaycan kültürüyle, folkloruyla, sosyal ve siyasi hayatıyla, ekonomik şartlarıyla ilgili önemli bilgiler aktarmaktadır. Ayrıca başta Baku olmak üzere Azerbaycan coğrafyasına ve çevre bölgelere ilişkin izlenimlerini öyküleştirmektedir. Bu makalede, Cihan Aktaş’ın Azize’nin Son Günü adlı öykü kitabına yansıyan Azerbaycan ve Azeri Türkü imgesi/algısı incelenmeye çalışılacaktır.Cihan Aktaş who is known by his books of thought and review as well as short stories published nine short story books between 1992 and 2009. The author, who refers to Azerbaijan and Azerbaijanis, especially in his short story book titled the Last Day of Azize (1997), either brought interesting portraits of Azerbaijani Turks in these texts or conveyed important information related to the culture, folklore, social and political life and economical conditions of Azerbaijan. He also narrated his impressions related to Azerbaijan geography and environmental area, especially to Baku. In this article, we will attempt to understand image/perception of Azerbaijan and Azerbaijanis reflecting on the book of Cihan Aktaş titled “the Last Day of Azize”
Emergency blood donor search system by Moora method
No full textKan, insanın yaşamsal fonksiyonlarının devam etmesinde hayati bir öneme sahiptir. Kaza, ameliyat ve hastalık gibi durumlarda hastanın acil kana ihtiyacı olabilmektedir. Bu gibi acil durumlar nedeniyle, ilgili kan grubuna sahip bireylere ulaşabilmek için günümüzde radyo, televizyon, web siteleri gibi araçlar kullanılmaktadır. Ancak bu araçlar aracılığıyla uygun donöre ulaşmak uzun zaman alabilmektedir. Acil durumlarda zaman oldukça önemlidir. Bu nedenle uygun kan grubuna sahip ve sağlık kuruluşuna yakın olan kişilere, hızlı şekilde ulaşılması gerekmektedir. Bu tezde, acil durumlarda sağlık kuruluşuna yakın ve uygun donörlere hızlı bir şekilde ulaşılarak kan verecekleri yerlere yönlendirilmesi hedeflenmiştir. Kan grubu ve donörlerin sağlık kuruluşuna uzaklık mesafesi, sağlık kurumu tarafından belirlenmekte ve acil kan duyurusu, kuruluşun belirlediği ölçütlere uyan tüm donörlere bildirim olarak gönderilmektedir. Bu çalışmada, sağlık kuruluşları duyuru girerek donörlere bildirim olarak ulaşabilmektedir. Çalışmanın sahip olduğu bu nitelik, literatüre sunduğu katkıları arttırdığı gibi konuya ilişkin benzer çalışmalara da öncü olacak nitelikleri taşımaktadır. Çalışma kapsamında bağışçı seçiminin gerçekleştirilmesinde Moora yönteminden yararlanılmıştır ve donörün sağlık kuruluşuna olan uzaklığı kadar diğer kriterlerin de seçimi etkilediği görülmüştür. Bu çalışma ile birlikte sağlık kuruluşları, hastalar ve yakınlarına da kazanımlar sağlanması amaçlanmaktadır.Blood is of great importance in maintaining human vital functions. In some cases, the patient should be given blood quickly. Because of this urgent situation, access to people with blood groups, radio, television, websites and so on such tools can be made through. In the time of death – survival race, people who can donate blood, even those with a suitable blood group and close to the health care provider, must be reached quickly. In this thesis, it is aimed to reach the appropriate donors close to the health institution quickly and be directed to the places where they will give blood. Emergency blood announcement, blood group and donors' possible distance to the hospital is determined by the health institution and sent as a notification to all donors who meet the criteria set by the institution. In this study, health organizations can access donors as notifications by entering a blood donation announcement. This quality of the study is expected to increase the contributions it makes to the literature and to lead similar studies on the subject. In the scope of the study, Moora method was used in the realization of donor selection. Moora method is chosen in the selection of donors in the study and it was found that other criteria affect the selection as well as the distance of the donor to the health institution. The aim of this study is to provide benefits to health institutions, patients and their relatives
Estimation of Exterior Parameters of Sensor Combination for Efficient 3D Modeling
No full textISPRS-Commission V Mid-Term Symposium on Close Range Image Measurement Technique -- JUN 21-24, 2010 -- Newcastle upon Tyne, ENGLANDLaser scanning is the latest method to collect 3D spatial data and modelling. It becomes more and more popular in engineering applications such as 3D modeling, deformation analysis, reverse engineering and urban planning. However, laser scanners are not sufficient on their own for 3D surveying and modeling. It should be used together with the photogrammetric data especially in the orientation of point clouds, measurement of the non-scanning areas, and texture mapping. This task needs the combination of range sensor and image sensor. The sensor combination like this is a very efficient and flexible system for close-range measurement applications. Most of the laser scanners include integrated camera but most of them are low resolution and they can not be removed independent from the scanner. Therefore high resolution camera should be mounted to the laser scanner. In this study we consider mounting a camera onto laser scanner and estimating exterior parameters with strong control methods. We also made experimental studies by using exterior parameters.ISPRS CommissScientific and Technological Research Council of Turkey ( TUBITAK)Turkiye Bilimsel ve Teknolojik Arastirma Kurumu (TUBITAK); "Foreign Research Scholarship Program" to Cihan ALTUNTAS; Vienna University of Technology; Scientific Research Found of The Selcuk University ( BAP)We would like to thank The Scientific and Technological Research Council of Turkey ( TUBITAK) for the financial support they have provided as a part of "Foreign Research Scholarship Program" to Cihan ALTUNTAS to the carried out this study in Vienna University of Technology. In addition author Cihan ALTUNTAS thanks Scientific Research Found of The Selcuk University ( BAP) to financial support to attend this scientific event
MAKİNE ÖĞRENMESİ VE AÇIKLANABİLİR YAPAY ZEKA (XAI) İLE SİGORTA SAHTEKÂRLIĞININ TESPİTİ
No full textThis thesis studies insurance fraud detection with machine learning (ML) methods
and explainable artificial intelligence (XAI) tools. The dataset contains insurance
policies and their related claims, where fraud is marked as a binary variable. The
preparation stage covers missing data handling, encoding of categorical fields, and
several feature engineering steps. Since fraud cases are rare, the Synthetic Minority
Oversampling Technique (SMOTE) is employed, and models are trained on both
the original and the balanced data. The algorithms tested include Logistic Regression,
Random Forest, Support Vector Machines (SVM), k-Nearest Neighbors (KNN),
LightGBM, and XGBoost. For XGBoost, two variants are considered: one with manually
designed features such as seasonality, polynomial terms, and interactions, and
another with automated feature engineering tuned by cross-validated AUC. Model
performance is evaluated with ROC-AUC, accuracy, recall, specificity, and balanced
accuracy, with thresholds optimized via ROC analysis. In addition, DeLong tests
are applied to assess whether the observed differences in AUC across models are
statistically significant. For interpretability, SHAP (SHapley Additive exPlanations)
is utilized. The results indicate that liability coverage and fault assignment are the
strongest drivers of fraud detection. Timing and policyholder features, such as claim
month and address changes, also contribute meaningfully, though at a lower level.
The findings showed that boosting methods achieved the best overall performance.
LightGBM and XGBoost outperformed the other models, with XGBoost slightly
ahead on the SMOTE-balanced data (AUC ≈ 0.85, recall ≈ 90%). Both manual and
automated feature engineering improved recall, thereby shifting the trade-off toward
catching more fraud cases. These findings emphasized the importance of designing
fraud detection models that combined strong predictive accuracy with transparent and
interpretable decision-making.Bu tezde, sigorta sahtekârlığının tespiti makine öğrenmesi (ML) yöntemleri ve açıklanabilir yapay zekâ (XAI) araçlarıyla incelenir. Veri seti, sigorta poliçeleri ve bunlara bağlı taleplerden oluşur ve sahtekârlık ikili bir değişken olarak tanımlanır. Ön işleme süreci eksik verilerin giderilmesini, kategorik alanların kodlanmasını ve çeşitli öznitelik mühendisliği adımlarını kapsar. Sahtekârlık vakalarının azlığı nedeniyle SMOTE (Synthetic Minority Oversampling Technique) yöntemi uygulanır ve modeller hem orijinal hem de dengelenmiş veriler üzerinde eğitilir. Test edilen algoritmalar arasında Lojistik Regresyon, Rassal Orman, Destek Vektör Makineleri (SVM), En Yakın Komşular (KNN), LightGBM ve XGBoost yer alır. XGBoost için iki varyant değerlendirilir: ilki mevsimsellik, polinom terimleri ve etkileşimler gibi elle tasarlanmış özniteliklere dayanır; ikincisi ise çapraz doğrulama AUC’sine göre ayarlanmış otomatik öznitelik mühendisliğini içerir. Model performansı ROC-AUC, doğruluk, duyarlılık, özgüllük ve dengeli doğruluk ölçütleri üzerinden, ROC analiziyle optimize edilen eşikler kullanılarak değerlendirilir. Ayrıca, modeller arasındaki AUC farklarının istatistiksel olarak anlamlı olup olmadığını test etmek için DeLong testleri uygulanır. Açıklanabilirlik için SHAP (SHapley Additive exPlanations) yöntemi kullanılır. Sonuçlar, sorumluluk kapsamı ve kusur durumunun sahtekârlık tespitinde en belirleyici unsurlar olduğunu gösteriyor. Talep ayı ve adres değişikliği gibi zamanlama ve poliçe sahibine ilişkin değişkenler de daha düşük düzeyde ancak anlamlı katkılar sağlıyor. Bulgular, boosting yöntemlerinin en iyi genel performansı elde ettiğini ortaya koyuyor. LightGBM ve XGBoost diğer modelleri geride bırakıyor; özellikle SMOTE ile dengelenmiş verilerde XGBoost biraz daha öne çıkıyor (AUC ≈0.85, duyarlılık ≈%90). Hem manuel hem de otomatik öznitelik mühendisliği, hatırlama oranını artırarak daha fazla sahtekârlık vakasının yakalanmasına katkı sağlıyor. Bu bulgular, güçlü tahmin performansıyla birlikte kararların nasıl verildiğini de şeffaf biçimde ortaya koyan sahtekârlık tespit modellerinin geliştirilmesinin önemini vurguluyor.M.S. - Master of Scienc
Insurance fraud detection via machine learning and explainable artificial intelligence (XAI)
No full textPhase transitions, elastic and electronic properties of hydrogen storage Na2PdH4
No full textHydrogen can be absorbed by some materials at specific pressures and temperatures. This is extremely important in terms of creating carbon-free and sustainable society. In this work hydrides are good candidates to fulfill these aims. Electronic and magnetic properties of hydrides are investigated. This study considers Na2PdH4 as solid storage of hydrogen material. Hydrogen is highly soluble in palladium and can be stored in large amounts at ambient conditions. The structural evolution, electronic and elastic properties of Na2PdH4 has been investigated by means of density functional theory. The SIESTA software package is used with the generalized gradient approximation for the exchange-correlation functional and norm-conserving Troullier-Martins pseudopotentials. High-pressure computations have been carried out to reveal phase transitions. Na2PdH4 is transformed from I4/mmm tetragonal structure to Immm orthorhombic structure at 100 GPa. The electronic band structures and density of states are obtained for both phases. Mechanical stability is analyzed using the elastic constants. Moreover, several parameters such as Young's modulus, shear modulus, and their ratios are obtained and discussed. © 2021 Author(s)
Decentralized multi-agent path finding framework and strategies based on automated negotiation
No full textThis paper introduces a negotiation framework to solve the Multi-Agent Path Finding (MAPF) Problem for self-interested agents in a decentralized fashion. The framework aims to achieve a good trade-off between the privacy of the agents and the effectiveness of solutions. Accordingly, a token-based bilateral negotiation protocol and two negotiation strategies are presented. The experimental results over four different settings of the MAPF problem show that the proposed approach could find conflict-free path solutions albeit suboptimally, especially when the search space is large and high-density. In contrast, Explicit Estimation Conflict-Based Search (EECBS) struggles to find optimal solutions. Besides, deploying a sophisticated negotiation strategy that utilizes information about local density for generating alternative paths can yield remarkably better solution performance in this negotiation framework.Interactive Intelligenc
A Pedagogical Stylistic Study of “Eleven” by Sandra Cisneros
No full textKnowing about the linguistic and stylistic features of any literary texts in general and the aspect of lexical pattern in particular, concerning the way through which the author uses adjectives, verbs, nouns and adverbs provide us with precious information that can be helpful in judging about the appropriateness of the text to be used for a specific purpose such as teaching to EFL students in a specific level. Therefore, according to the importance of this issue, and based on the checklist of linguistic and stylistic categories proposed by Leech and Short (2000), the stylistic analysis of the short story entitled “Eleven” written by Sandra Cisneros was performed in this study from the lexical category dimension. In the end, the results and findings of the study are elaborated and discussed, with the conclusion that the author was successful in using the appropriate diction and linguistic features to express her main points from the tongue of an Eleven-year-old kid who is also ten, nine, eight, seven, six, five, four, two and one year old.
Key words: Lexical category, Eleven, Stylistic analysis, EFL, Short stor
- …
