No abstract

No abstract

No abstract

Aims: The interruption of blood flow to the heart muscle is called a heart attack. During a heart attack, the risk of permanent damage increases with every second the heart tissue cannot receive enough blood. If early and appropriate intervention is not performed, loss of heart tissue occurs. Causes such as smoking, cholesterol, diabetes, high blood pressure, old age, obesity, genetics, and high levels of certain substances produced in the liver are the main risk factors for heart attack. This study aims to predict the risk of heart attack with machine learning methods using a dataset created by considering risk factors. Methods: The performances of three types of Linear Discriminant Analysis classifiers, Normal, Ledoit-Wolf, and Oracle Shrinkage Approximating, were compared on the Cleveland dataset. Results: Normal Linear Discriminant Analysis made the best classification with 83.60% accuracy and performed better than regularized versions. Conclusion: Linear Discriminant Analysis methods are a promising classifier for heart attack prediction and can be applied in hospitals as an objective and automated system that eases specialists' workload and helps reduce diagnostic costs.

As the number of government and commercial satellites increases, there is a large increase in Earth observation (EO) imagery. Using different locations and tools, images can be taken from more than one satellite. Manipulations are carried out on these images using a variety of different methods. The number of studies that have been done on the manipulation of EO images is very small. In recent years, generative adversarial networks (GANs), a major breakthrough in deep learning, have made it very easy to obtain fake images. In this study, scene-by-scene fake images were obtained with the deep convolutional GAN on the EuroSAT dataset, which is one of the EO image sets, and fake scene images were obtained from the original scenes. In this study, a dataset called RF-EuroSAT was created. It consists of 14 classes and 36,000 images. Five transfer learning models (VGG-16, DenseNet201, MobileNetV2, RegNetY320, and ResNet152V2) were used to classify this dataset. Using these models as feature extraction and ensemble models (XGBoost, CatBoost, and LightGBM) as classifiers, the classification process was performed using our proprietary transferemble model. The best result was obtained with an accuracy of 91.55% using our transferemble model, which is developed in a modular structure.