The significant role of Li-ion batteries (LIBs) in electric vehicles (EVs) emphasizes their advantages in terms of energy density, being lightweight, and being environmentally sustainable. Despite their obstacles, such as costs, safety concerns, and recycling challenges, LIBs are crucial in terms of the popularity of EVs. The accurate prediction and management of LIBs in EVs are essential, and machine learning-based methods have been explored in order to estimate parameters such as the state of charge (SoC), the state of health (SoH), and the state of power (SoP). Various machine learning techniques, including support vector machines, decision trees, and deep learning, have been employed for predicting LIB states. This study proposes a methodology for comparative analysis, focusing on classical and deep learning approaches, and discusses enhancements to the LSTM (long short-term memory) and Bi-LSTM (bidirectional long short-term memory) methods. Evaluation metrics such as MSE, MAE, RMSE, and R-squared are applied to assess the proposed methods’ performances. The study aims to contribute to technological advancements in the electric vehicle industry by predicting the performance of LIBs. The structure of the rest of the study is outlined, covering materials and methods, LIB data preparation, analysis, the proposal of machine learning models, evaluations, and concluding remarks, with recommendations for future studies.

Derin öğrenme makine öğreniminin bir koludur. Makine öğreniminin başlarından günümüze kadar geçen süreçte yapay zekaya olan ilgi giderek artmış ve günümüzde en çok kullanılan yapay zeka algoritmaları olan derin öğrenme mimarilerinin ortaya çıkmasını sağlamıştır. Derin öğrenme mimarileri ile birlikte yapay zeka problemlerinin çözümü için pek çok derin öğrenme yaklaşımları geliştirilmiştir. Endüstri, tıp, robotik, görüntü işleme, bilgisayar görmesi, nesne tespiti, ses işleme-tanıma, çeviri, gelecek tahmini, finansal gibi pek çok alanda akıllı çözümler üretmektedir. Bu çalışmada, derin öğrenme mimarileri ve algoritmaları incelenerek, literatürde yapılmış çalışmalar ışığında uygulama alanları temelinde başarımları değerlendirilmiştir. Derin öğrenme mimarileri ile birlikte derin öğrenmede kullanılan kütüphanelere yer verilmiştir. Bununla beraber farklı problemlerin çözümlerine yönelik geliştirilen derin öğrenme mimarileri yer almaktadır.

This research presents a Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) model developed for malware classification from IoT devices in the SCADA system and for detecting anomalies in the network. The developed model identifies complex attacks in the network by taking advantage of the strengths of CNNs that reveal spatial features and LSTMs that detect temporal dependency. CICIoT 2023 is used as the dataset. ADAM optimization algorithm with cross-entropy loss is used to eliminate overfitting and training is performed. Within the scope of the study, the proposed model is compared by applying it with six deep learning architectures (hybrid CNN-LSTM, Non-Local Neural Network (NLNN), Residual Attention Network (RAN), Dual Attention Network (DANet), Transformer-CNN and Attentional CNN). The obtained results show that the proposed CNN-LSTM model outperforms other complex architectures and achieves a high test accuracy of 99.23%. It has demonstrated remarkable performance according to precision, recall and F1 evaluation metrics in detecting distributed denial of service (DDoS) and denial of service (DoS) attacks. The proposed model successfully identifies Mirai botnet variants and fragmentation-based attacks. Although other models, Transformer-CNN (98.81%) and DANet (98.07%), provide high performance, they fall behind the superior temporal modeling capabilities of CNN-LSTM. When the obtained findings are examined, they highlight the relative strengths of various deep learning approaches for IoT security applications. The performances of the implemented deep learning models reached accuracies exceeding 96%, demonstrating the importance of SCADA-based IoT systems against evolving cyber threats. The study revealed the superior successes of deep learning-based approaches for IoT security.

Classification of skin diseases is a important isssue for early diagnosis and treatment. The process of determining the disease by the specialist physician also delays the treatment process to be applied to the patient. Computer-aided diagnosis systems play an important role in early diagnosis and initiation of treatment by minimizing such processes. In this study, high-performance classification of skin lesions was performed by using Deep Learning models. Dataset was ISIC data set, dataset were expanded by using data augmentation techniques. In the images in this dataset, there are images of Actinic Keratosis, Dermatofibroma, Pigmented Benign Keratosis, Seborrheic Keratosis, Vascular Lesion skin diseases. The data set was classified by Deep Learning models by using the supervised learning method.. SequeezeNet, AlexNet, GoogleNet, Vgg-19, ResNet101, DenseNet201, ResNet-50, ResNet-18, Vgg-16 DL models were used for classification. To evaluate of classification success of Deep Learning models, confusion matrix and F1-score, precision, sensitivity and accuracy metrics obtained from the matrix were used. According to the F1-score, the most successful model is Vgg16 with 97.41%, while the highest accuracy rate obtained by ResNet18 with 98.06%. High success rate shows that such systems can be used for diagnosis and treatment processes.