10 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 10 / 10
Yayın Sınıflandırma için diferansiyel mahremiyete dayalı öznitelik seçimi(Gazi Univ, Fac Engineering Architecture, 2018) Var, Esra; İnan, AliVeri madenciliği ve makine öğrenmesi çözümlerinin en önemli ön aşamalarından biri yapılacak analizde kullanılacak verinin özniteliklerinin uygun bir alt kümesini belirlemektir. Sınıflandırma yöntemleri için bu işlem, bir özniteliğin sınıf niteliği ile ne oranda ilişkili olduğuna bakılarak yapılır. Kişisel gizliliği koruyan pek çok sınıflandırma çözümü bulunmaktadır. Ancak bu yöntemler için öznitelik seçimi yapan çözümler geliştirilmemiştir. Bu çalışmada, istatistiksel veritabanı güvenliğinde bilinen en kapsamlı ve güvenli çözüm olan diferansiyel mahremiyete dayalı özgün öznitelik seçimi yöntemleri sunulmaktadır. Önerilen bu yöntemler, yaygın olarak kullanılan bir veri madenciliği kütüphanesi olan WEKA ile entegre edilmiş ve deney sonuçları ile önerilen çözümlerin sınıflandırma başarımına olumlu etkileri gösterilmiştir.Yayın ICamp - The educational web for higher education(Springer Verlag, 2006) Kieslinger, Barbara; Wild, Fridolin; Arsun, Onur İhsaniCamp is an EC-funded research project in the area of Technology Enhanced Learning (TEL) that aims to support collaboration and social networking across systems, countries and disciplines in higher education. The concept of an iCamp Space will build on existing interfaces and integrate shared community features. Interoperability amongst different open source learning systems and tools is the key to successful sustainability of iCamp. The content for this collaboration within social communities is provided via distributed networked repositories including, for example, content brokerage platforms, online libraries, and learning object databases. The innovative pedagogical model of iCamp is based on social constructivist learning theories. iCamp creates an environment for a new way of social networking in higher education that puts more emphasis on self-organised, self-directed learning, social networking and cross-cultural collaboration.Yayın Effective semi-supervised learning strategies for automatic sentence segmentation(Elsevier Science BV, 2018-04-01) Dalva, Doğan; Güz, Ümit; Gürkan, HakanThe primary objective of sentence segmentation process is to determine the sentence boundaries of a stream of words output by the automatic speech recognizers. Statistical methods developed for sentence segmentation requires a significant amount of labeled data which is time-consuming, labor intensive and expensive. In this work, we propose new multi-view semi-supervised learning strategies for sentence boundary classification problem using lexical, prosodic, and morphological information. The aim is to find effective semi-supervised machine learning strategies when only small sets of sentence boundary labeled data are available. We primarily investigate two semi-supervised learning approaches, called self-training and co-training. Different example selection strategies were also used for co-training, namely, agreement, disagreement and self-combined. Furthermore, we propose three-view and committee-based algorithms incorporating with agreement, disagreement and self-combined strategies using three disjoint feature sets. We present comparative results of different learning strategies on the sentence segmentation task. The experimental results show that the sentence segmentation performance can be highly improved using multi-view learning strategies that we proposed since data sets can be represented by three redundantly sufficient and disjoint feature sets. We show that the proposed strategies substantially improve the average baseline F-measure of 67.66% to 75.15% and 64.84% to 66.32% when only a small set of manually labeled data is available for Turkish and English spoken languages, respectively.Yayın Incremental construction of classifier and discriminant ensembles(Elsevier Science Inc, 2009-04-15) Ulaş, Aydın; Semerci, Murat; Yıldız, Olcay Taner; Alpaydın, Ahmet İbrahim EthemWe discuss approaches to incrementally construct an ensemble. The first constructs an ensemble of classifiers choosing a subset from a larger set, and the second constructs an ensemble of discriminants, where a classifier is used for some classes only. We investigate criteria including accuracy, significant improvement, diversity, correlation, and the role of search direction. For discriminant ensembles, we test subset selection and trees. Fusion is by voting or by a linear model. Using 14 classifiers on 38 data sets. incremental search finds small, accurate ensembles in polynomial time. The discriminant ensemble uses a subset of discriminants and is simpler, interpretable, and accurate. We see that an incremental ensemble has higher accuracy than bagging and random subspace method; and it has a comparable accuracy to AdaBoost. but fewer classifiers.Yayın Mapping classifiers and datasets(Pergamon-Elsevier Science Ltd, 2011-04) Yıldız, Olcay TanerGiven the posterior probability estimates of 14 classifiers on 38 datasets, we plot two-dimensional maps of classifiers and datasets using principal component analysis (PCA) and Isomap. The similarity between classifiers indicate correlation (or diversity) between them and can be used in deciding whether to include both in an ensemble. Similarly, datasets which are too similar need not both be used in a general comparison experiment. The results show that (i) most of the datasets (approximately two third) we used are similar to each other, (ii) multilayer perceptrons and k-nearest neighbor variants are more similar to each other than support vector machine and decision tree variants. (iii) the number of classes and the sample size has an effect on similarity.Yayın Adaptive convolution kernel for artificial neural networks(Academic Press Inc., 2021-02) Tek, Faik Boray; Çam, İlker; Karlı, DenizMany deep neural networks are built by using stacked convolutional layers of fixed and single size (often 3 × 3) kernels. This paper describes a method for learning the size of convolutional kernels to provide varying size kernels in a single layer. The method utilizes a differentiable, and therefore backpropagation-trainable Gaussian envelope which can grow or shrink in a base grid. Our experiments compared the proposed adaptive layers to ordinary convolution layers in a simple two-layer network, a deeper residual network, and a U-Net architecture. The results in the popular image classification datasets such as MNIST, MNIST-CLUTTERED, CIFAR-10, Fashion, and ‘‘Faces in the Wild’’ showed that the adaptive kernels can provide statistically significant improvements on ordinary convolution kernels. A segmentation experiment in the Oxford-Pets dataset demonstrated that replacing ordinary convolution layers in a U-shaped network with 7 × 7 adaptive layers can improve its learning performance and ability to generalize.Yayın Closeness and uncertainty aware adversarial examples detection in adversarial machine learning(Elsevier Ltd, 2022-07) Tuna, Ömer Faruk; Çatak, Ferhat Özgür; Eskil, Mustafa TanerWhile deep learning models are thought to be resistant to random perturbations, it has been demonstrated that these architectures are vulnerable to deliberately crafted perturbations, albeit being quasi-imperceptible. These vulnerabilities make it challenging to deploy Deep Neural Network (DNN) models in security-critical areas. Recently, many research studies have been conducted to develop defense techniques enabling more robust models. In this paper, we target detecting adversarial samples by differentiating them from their clean equivalents. We investigate various metrics for detecting adversarial samples. We first leverage moment-based predictive uncertainty estimates of DNN classifiers derived through Monte-Carlo (MC) Dropout Sampling. We also introduce a new method that operates in the subspace of deep features obtained by the model. We verified the effectiveness of our approach on different datasets. Our experiments show that these approaches complement each other, and combined usage of all metrics yields 99 % ROC-AUC adversarial detection score for well-known attack algorithms.Yayın Sarcasm detection on news headlines using transformers(Springer, 2025-09-07) Gümüşçekiçci, Gizem; Dehkharghani, RahimSarcasm poses a linguistic challenge due to its figurative nature, where intended meaning contradicts literal interpretation. Sarcasm is prevalent in human communication, affecting interactions in literature, social media, news, e-commerce, etc. Identifying the true intent behind sarcasm is challenging but essential for applications in sentiment analysis. Detecting sarcasm in written text, as a challenging task, has attracted many researchers in recent years. This paper attempts to detect sarcasm in news headlines. Journalists prefer using sarcastic news headlines as they seem much more interesting to the readers. In the proposed methodology, we experimented with Transformers, namely the BERT model, and several Machine and Deep Learning models with different word and sentence embedding methods. The proposed approach inherently requires high-performance resources due to the use of large-scale pre-trained language models such as BERT. We also extended an existing news headlines dataset for sarcasm detection using augmentation techniques and annotating it with hand-crafted features. The proposed methodology could outperform almost all existing sarcasm detection approaches with a 98.86% F1-score when applied to the extended news headlines dataset, which we made publicly available on GitHub.Yayın ANN activation function estimators for homomorphic encrypted inference(Institute of Electrical and Electronics Engineers Inc., 2025-06-13) Harb, Mhd Raja Abou; Çeliktaş, BarışHomomorphic Encryption (HE) enables secure computations on encrypted data, facilitating machine learning inference in sensitive environments such as healthcare and finance. However, efficiently handling non-linear activation functions, specifically Sigmoid and Tanh, remains a significant computational challenge for encrypted inference using Artificial Neural Networks (ANNs). This study introduces a lightweight, ANN-based estimator designed to accurately approximate activation functions under homomorphic encryption. Unlike traditional polynomial and piecewise linear approximations, the proposed ANN estimators achieve superior accuracy with lower computational overhead associated with bootstrapping or high-degree polynomial techniques. These estimators are trained on plaintext data and seamlessly integrated into encrypted inference pipelines, significantly outperforming conventional methods. Experimental evaluations demonstrate notable improvements, with ANN estimators enhancing accuracy by approximately 2% for Sigmoid and up to 73% for Tanh functions, improving F1-scores by approximately 2% for Sigmoid and up to 88% for Tanh, and markedly reducing Mean Square Error (MSE) by up to 96% compared to polynomial approximations. The ANN estimator achieves an accuracy of 97.70% and an AUC of 0.9997 when integrated into a CNN architecture on the MNIST dataset, and an accuracy of 85.25% with an AUC of 0.9459 on the UCI Heart Disease dataset during ciphertext inference. These results underscore the estimator’s practical effectiveness and computational feasibility, making it suitable for secure and efficient ANN inference in encrypted environments.Yayın Designing a scalable agricultural information system for pest detection and decision support in hazelnut cultivation(World Scientific Publishing Company, 2025-11-12) Aydın, ŞahinThis study presents a microservices-based, multi-tiered information system to detect, monitör and manage pest species that cause yield losses in hazelnut production. The system integrates a deep learning model for classifying pest images submitted by field users, the generation of pest density maps and location-based early warning mechanisms for growers. Delivered through mobile, web and desktop platforms, the system enables data sharing among farmers, researchers and decision-makers, supporting agricultural decisions. Experimental findings show that the DNN+ResNet50 architecture achieved the highest accuracy (91.88%) among all tested CNN models. Performance evaluations indicated that the Authentication and Heatmap services sustained high stability under loads of up to 1000 requests, while the Bug Classification Service was reliable up to 750 requests before reaching a critical resource threshold. The usability test resulted in an overall score of 38 out of 50, with sub-scores of Appropriateness Recognizability (0.73, Acceptable), Learnability (0.71, Acceptable), Operability (0.65, Questionable), User Error Protection (0.86, Good), User Interface Aesthetics (0.83, Good) and Accessibility (0.74, Acceptable). With its robust technical architecture and practical implementation, the proposed system can generate economic, social and commercial outcomes. This study provides a software engineering-oriented approach to the digitalization of agricultural production and the sustainable management of pests.
