Araştırma Çıktıları | WoS | Scopus | TR-Dizin | PubMed
Permanent URI for this communityhttps://hdl.handle.net/20.500.14719/1741
Browse
5 results
Search Results
Publication Metadata only The analysis of feature selection with machine learning for indoor positioning(Institute of Electrical and Electronics Engineers Inc., 2021) Aydin, Hurkan M.; Ali, Muhammad Ammar; Gelal, Ece; Aydin, Hurkan M., Department of Electrical and Electronic Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Ali, Muhammad Ammar, Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Gelal, Ece, Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyIndoor positioning is useful in various venues including warehouses, convention centers, malls, airports, nursing homes. In these scenarios, reducing the complexity of location estimation both improves responsiveness and helps to elongate battery life of the mobile device. In this work, we carry out a detailed analysis of the impact of Principal Component Analysis (PCA) on the computational complexity and accuracy with different machine learning algorithms on a large data set containing 520 APs. We compare the algorithms' training and testing times, as well as their accuracies in the presence and absence of PCA. Our results show that (i) PCA significantly reduces both the training and testing times for classification and regression using k-nearest neighbor (kNN) and support vector machine (SVM) algorithms while preserving if not improving accuracy, (ii) PCA slightly improves the training/testing times for regression using multi-layer perceptron (MLP), (iii) random forest (RF) does not perform well with PCA. © 2021 Elsevier B.V., All rights reserved.Publication Metadata only Evaluation of the Effects of Noise and Sampling Rate on Detection of High Impedance Fault with Machine Learning Methods on the Distribution System(Institute of Electrical and Electronics Engineers Inc., 2023) Baharozu, Eren; Ilhan, Suat; Soykan, Gurkan; Sandikkaya, M.T.; Usta, O.; Baharozu, Eren, Department of Electrical Engineering, İstanbul Teknik Üniversitesi, Istanbul, Turkey; Ilhan, Suat, Department of Electrical Engineering, İstanbul Teknik Üniversitesi, Istanbul, Turkey; Soykan, Gurkan, Department of Energy Systems Engineering, Bahçeşehir Üniversitesi, Istanbul, TurkeyHigh impedance faults (HIF) are complex faults that affect the stability and reliability of the distribution networks. In the last decades, machine learning (ML) methods have been used extensively to detect such faults. However, their performances were not evaluated completely in the studies. Thus, an analysis of different ML methods, which are artificial neural network (ANN), support vector machine (SVM), and k-nearest neighbors (KNN) was presented in this paper for HIF identification. The precision of methods was tested and compared for high impedance fault detection by considering the effects of noise and sampling rate. Moreover, a combination of the discrete wavelet transform (DWT) with machine learning methods was investigated. A comparison was made with the accuracy rate of algorithms in distinguishing HIFs from other events. IEEE 34 busses test network was modeled in Matlab/Simulink, and current waveforms are utilized for training the algorithms. Based on the results, it is revealed that the accuracy of all algorithms was decreased with lower sampling rates and higher noise ratios in the signal. In terms of the comparison of methods, KNN became the most endurance algorithm to change in sampling rate, while ANN was the least influenced by noise. Additionally, SVM showed better precisions when noise is not added to the signal. © 2023 Elsevier B.V., All rights reserved.Publication Open Access Assessing Dyslexia with Machine Learning: A Pilot Study Utilizing Google ML Kit(Institute of Electrical and Electronics Engineers Inc., 2023) Eroğlu, Günet; Harb, Mhd Raja Abou; Eroğlu, Günet, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Harb, Mhd Raja Abou, Department of Computer Engineering, Işik Üniversitesi, Istanbul, TurkeyIn this study, we explore the application of Google ML Kit, a machine learning development kit, for dyslexia detection in the Turkish language. We collected face-tracking data from two groups: 49 dyslexic children and 22 typically developing children. Using Google ML Kit and other machine learning algorithms based on eye-tracking data, we compared their performance in dyslexia detection. Our findings reveal that Google ML Kit achieved the highest accuracy among the tested methods. This study underscores the potential of machine learning-based dyslexia detection and its practicality in academic and clinical settings. © 2024 Elsevier B.V., All rights reserved.Publication Metadata only Triplet MAML for Few-Shot Classification Problems(Springer Science and Business Media Deutschland GmbH, 2024) Gülcü, Ayla; Özkan, İsmail Taha Samed; Kus, Zeki; Karakuş, Osman Furkan; Ortis, A.; Hameed, A.A.; Jamil, A.; Gülcü, Ayla, Department of Software Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Özkan, İsmail Taha Samed, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Kus, Zeki, Department of Computer Engineering, Bahçeşehir Üniversitesi, Istanbul, Turkey; Karakuş, Osman Furkan, Department of Computer Engineering, Fatih Sultan Mehmet Vakıf Üniversitesi, Istanbul, Turkey, Department of Computer Engineering, Yıldız Teknik Üniversitesi, Istanbul, TurkeyIn this study, we propose a TripletMAML algorithm as an extension to Model-Agnostic Meta-Learning (MAML) which is the most widely-used optimization-based meta-learning algorithm. We approach MAML from a metric-learning perspective and train it using meta-learning tasks composed of triplets of images. The idea of meta-learning is preserved while generating the meta-learning tasks and training our novel meta-model. The experimental results obtained on four few-shot classification datasets show that TripletMAML that is trained using a combined loss yields in high quality results. We compared the performance of TripletMAML to several metric learning-based methods and a baseline method, in addition to MAML. For fair comparison, we used the reported results of those algorithms that were obtained using the same shallow backbone. The results show that TripletMAML improves MAML by a large margin, and yields better results than most of the compared algorithms in both 1-shot and 5-shot settings. Moreover, when we consider the classification performance of other meta-learning algorithms that use much deeper backbones, we conclude that TripletMAML is not only competitive in terms of the classification performance but also very efficient in terms of the complexity. © 2024 Elsevier B.V., All rights reserved.Publication Metadata only Image Processing and Machine Learning Techniques for Chagas Disease Detection and Identification(Springer Science and Business Media Deutschland GmbH, 2024) Rada, Lavdie; Azzawi, Inass; Kumar, Preet; Brito-Loeza, Carlos; Karabaǧ, Cefa; Reyes-Aldasoro, Constantino Carlos; Yap, M.H.; Kendrick, C.; Behera, A.; Cootes, T.; Zwiggelaar, R.; Rada, Lavdie, Faculty of Engineering and Natural Sciences, Bahçeşehir Üniversitesi, Istanbul, Turkey; Azzawi, Inass, Faculty of Engineering and Natural Sciences, Bahçeşehir Üniversitesi, Istanbul, Turkey; Kumar, Preet, University of the West of England, Bristol, United Kingdom; Brito-Loeza, Carlos, Universidad Autónoma de Yucatán, Merida, Mexico; Karabaǧ, Cefa, School of Computing and Digital Media, London Metropolitan University, London, United Kingdom; Reyes-Aldasoro, Constantino Carlos, School of Science & Technology, University of London, London, United KingdomChagas disease, caused by the Trypanosoma cruzi parasite, poses a significant health threat, particularly in Latin America, with millions affected globally. This research introduces a novel approach using deep learning techniques for the automated detection of Trypanosoma cruzi in blood smear images provided by Zoonoses Laboratory (CIR) in Mexico. Advanced deep learning architectures like Faster RCNN, RetinaNet, YOLOv8, and FCOS have been adapted, trained, and compared with each other in terms of the detection accuracy of each image. Our selection of those models is based on their ability to swiftly and accurately detect anomalies, measured through rigorous assessment using pivotal metrics like Mean Average Precision (mAP) across varying Intersection over Union (IoU) thresholds. Notably, the YOLOv8 model has showcased outstanding performance, boasting a remarkable mAP score of 0.951 for parasite detection and localisation. Specifically, YOLOv8 outperforms with a leading mAP of 0.951 at 50% IoU and maintains commendable precision with a score of 0.594 for IoU thresholds ranging from 50% to 95%. This research reduces dependence on skilled manual analysis holding a significant implications for healthcare in Chagas-affected regions by providing a rapid, automated solution to disease detection. This work has the potential to revolutionise diagnostics in resource-limited settings. Moreover, the models’ adaptability to other parasitic infections enhances their global health impact. © 2024 Elsevier B.V., All rights reserved.