Yazar "Yilmaz, Atinc" seçeneğine göre listele

Listeleniyor 1 - 6 / 6
  • Küçük Resim Yok
    Öğe
    An Analysis of Dimension Reduction Methods Applicable for Out Of Sample Problem in Hyperspectral Images
    (IEEE, 2019) Yilmaz, Atinc; Ozturk, Umit
    Dimension reduction (DR) techniques provide advantages in terms of reducing computational time and data storage requirements. These methods are widely used for Hyper-spectral Imaging (HSI) data classification purposes. However, the need of extention to out of sample (OOS) data during the learning procedure is a main drawback for traditional manifold learning (ML) algorithms. There are a number of DR methods applicable for OOS problem. This study aims at focusing a comprehensive analysis of these DR methods with HSI data classification. In this context, we made experiments on HSI data by applying LPP, NPE, OLPP, ONPE and Nystrom (for LE and LLE) algorithms to observe classification performance. Firstly, parametric graphs are obtained to understand the behaviour and then, classification performance is measured upon reduced dimensionality. Support Vector Machines (SVM) and K-Nearest Neighbor (KNN) are used as classifier. All experiments are conducted on Kennedy Space Center (KSC) and Indian Pines hyperspectral data.
  • Küçük Resim Yok
    Öğe
    Hybrid Handwriting Character Recognition with Transfer Deep Learning
    (IEEE, 2019) Can, Ferit; Yilmaz, Atinc
    Handwriting character recognition is useful and important in terms of allowing correct recognition and interpretation of all characters such as handwritten letters, numbers and figures. Deep convolutional neural networks have been used frequently for computer vision in recent years due to high performance in image processing, feature extraction and classification. A lot of sample data, processing power and time are needed to train CNNs. Transfer learning enables us to obtain specific CNNs for the classes we want by minimizing these needs In this work, firstly, different CNN models are trained with transfer learning by using NIST19 dataset with handwritten characters, and then a hybrid model is created by evaluating the results of each CNN and revealing the best value. As a result of the experiment carried out on the test data set, it is observed that a performance increase of 1.1% is achieved with the created model.
  • Küçük Resim Yok
    Öğe
    PREDICTION AND CLASSIFICATION OF PRESSURE INJURIES BY DEEP LEARNING
    (Termedia Publishing House Ltd, 2021) Yilmaz, Atinc; Kizil, Hamiyet; Kaya, Umut; cakir, Ridvan; Demiral, Melek
    Pressure injuries are a serious medical problem that both negatively affects the patient's quality of life and results in significant healthcare costs. In cases where a patient doesn't receive appropriate treatment and care, death may result. Nurses play critical roles in the prevention, care, and treatment of pressure injuries as members of the healthcare team who closely monitor the health status of the patient. Today, the use of artificial intelligence is becoming more prevalent in healthcare, as in many other areas. Artificial intelligence is a method that aims to solve complex problems by using computers to mathematically simulate the way the brain works. In this article, we compile and share information about a deep learning model developed for the detection and classification of pressure injuries. Deep learning can operate on many types of data. Convolutional Neural Networks (CNN) prefer images because they can handle 2D arrays. In this case, the images, annotated according to the National Pressure Injury Advisory Panel pressure injury classification system, have been fed into a deep learning model using CNN. The developed CNN model has a 97% success in detecting and classifying pressure injuries, and as more images are collected and fed into the CNN, the prediction accuracy will increase. This deep learning model allows for the automatic detection and classification of pressure injuries, an indicator of health outcomes, at an early stage and for quick and accurate intervention. In this context, it is expected that the quality of nursing care will increase, the prevalence of pressure injury will decrease, and the economic burden of this health problem will decrease.
  • Küçük Resim Yok
    Öğe
    Risk analysis of lung cancer and effects of stress level on cancer risk through neuro-fuzzy model
    (Elsevier Ireland Ltd, 2016) Yilmaz, Atinc; Ari, Seckin; Kocabicak, Umit
    A significant number of people pass away due to limited medical resources for the battle with cancer. Fatal cases can be reduced by using the computational techniques in the medical and health system. If the cancer is diagnosed early, the chance of successful treatment increases. In this study, the risk of getting lung cancer will be obtained and patients will be provided with directions to exterminate the risk. After calculating the risk value for lung cancer, status of the patient's susceptibility and resistance to stress is used in determining the effects of stress to disease. In order to resolve the problem, the neuro-fuzzy logic model has been presented. When encouraging results are obtained from the study; the system will form a pre-diagnosis for the people who possibly can have risk of getting cancer due to working conditions or living standards. Therefore, this study will enable these people to take precautions to prevent the risk of cancer. In this study a new t-norm operator has been utilized in the problem. Finally, the performance of the proposed method has been compared to other methods. Beside this, the contribution of neuro-fuzzy logic model in the field of health and topics of artificial intelligence will also be examined in this study. (C) 2016 Elsevier Ireland Ltd. All rights reserved.
  • Küçük Resim Yok
    Öğe
    Sentiment Analysis of Elon Musk's Twitter Data Using LSTM and ANFIS-SVM
    (Springer International Publishing Ag, 2022) Erkartal, Bugra; Yilmaz, Atinc
    Social media plays a huge role spreading words to millions and influencing their opinions. Twitter is one of the most essential platform that reach over 300 million active users and 500 million tweets per day, it plays a significant role spreading the word around the world,. These tweets covers a various subjects from personal conversations to globally important topics such as updates about Covidl9 and macroeconomic subjects. Especially in financial matters, it is a very common situation that business owners, even politicians report the news on Twitter first. The Tesla's and SpaceX's CEO and owner Elon Musk's tweets had a huge impact on coin market or even stock exchanges. Although many accused him of market manipulation his tweets impact cannot be underestimated. In 2020 and 2021 there are various tweets that strike the stock market instantly both in the positive and negative direction. This study aims to predict the direction of his tweets and perform a sentiment analysis using both Long-Short Term Memory (LSTM) and Adaptive Neuro Fuzzy Interface Systems (ANFIS)-SVM(Support Vector Machines) models. The dataset is obtained by using Twitter API which spans a time horizon of 5 years. In order to compare the results under same conditions same preprocessing steps are performed for both models. According to the results, LSTM performs a superior performance with its 72.2% accuracy against ANFIS-SVM model with 74.1%.
  • Küçük Resim Yok
    Öğe
    Sepsis Prediction by Using a Hybrid Metaheuristic Algorithm: A Novel Approach for Optimizing Deep Neural Networks
    (Mdpi, 2023) Kaya, Umut; Yilmaz, Atinc; Asar, Sinan
    The early diagnosis of sepsis reduces the risk of the patient's death. Gradient-based algorithms are applied to the neural network models used in the estimation of sepsis in the literature. However, these algorithms become stuck at the local minimum in solution space. In recent years, swarm intelligence and an evolutionary approach have shown proper results. In this study, a novel hybrid metaheuristic algorithm was proposed for optimization with regard to the weights of the deep neural network and applied for the early diagnosis of sepsis. The proposed algorithm aims to reach the global minimum with a local search strategy capable of exploring and exploiting particles in Particle Swarm Optimization (PSO) and using the mental search operator of the Human Mental Search algorithm (HMS). The benchmark functions utilized to compare the performance of HMS, PSO, and HMS-PSO revealed that the proposed approach is more reliable, durable, and adjustable than other applied algorithms. HMS-PSO is integrated with a deep neural network (HMS-PSO-DNN). The study focused on predicting sepsis with HMS-PSO-DNN, utilizing a dataset of 640 patients aged 18 to 60. The HMS-PSO-DNN model gave a better mean squared error (MSE) result than other algorithms in terms of accuracy, robustness, and performance. We obtained the MSE value of 0.22 with 30 independent runs.