Browsing by Author "Canayaz, M."

Now showing 1 - 7 of 7

Crypto-Currency Sentiment Analyse on Social Media
(Institute of Electrical and Electronics Engineers Inc., 2019) Erdoǧan, M.C.; Canayaz, M.
The amount of data produced with the introduction of social media tools has reached gigantic size. In analyzing data, and now traditional methods are no longer enough, the concept of 'Big Data' has entered our lives. There is an inevitable need to produce meaningful summaries by analyzing data at huge size. Large data tools are used to meet this need. In order to have knowledge about human behaviors using these tools and to develop solutions in this direction, social media and especially Twitter data are being studied. As is known, the concept of crypto currency, which has come from trends in recent years, attracts more and more people. In this study, approaches to the most commonly used crypto currencies were examined for the first time using large data tools. The meaningless data on twits which is taken from Twitter has been cleared and the approaches of users against crypto paralysis have been analyzed by using various classifiers and the results have been shown. © 2018 IEEE.
Feature Selection With the Whale Optimization Algorithm and Artificial Neural Network
(Institute of Electrical and Electronics Engineers Inc., 2017) Canayaz, M.; Demir, M.
Feature selection is addressed an important problem in data mining. To be high dimension of the data obtained from the sources is encountered as an issue in many issues such as computation cost. For this reason, eliminating the unnecessary ones among these data and choosing the appropriate ones makes it possible to evaluate the information correctly. In this study, it is tried to suggest a method that can be used in feature selection on data sets. In this method, The Whale Optimization Algorithm, which is one of the new meta-heuristic algorithms, is used to select appropriate features. Training with artificial neural networks takes place during the evaluation process of selected features. At the end of the training, the features that provide the minimum error value are selected. In the performance evaluation of the method, known data sets will be used and the results will be given in comparison with the Particle Swarm Optimization method. © 2017 IEEE.
Machine Learning-Based Prediction of Length of Stay (Los) in the Neonatal Intensive Care Unit Using Ensemble Methods
(Springer Science and Business Media Deutschland GmbH, 2024) Erdogan Yildirim, A.; Canayaz, M.
Neonatal medical data holds critical information within the healthcare industry, and it is important to analyze this data effectively. Machine learning algorithms offer powerful tools for extracting meaningful insights from the medical data of neonates and improving treatment processes. Knowing the length of hospital stay in advance is very important for managing hospital resources, healthcare personnel, and costs. Thus, this study aims to estimate the length of stay for infants treated in the Neonatal Intensive Care Unit (NICU) using machine learning algorithms. Our study conducted a two-class prediction for long and short-term lengths of stay utilizing a unique dataset. Adopting a hybrid approach called Classifier Fusion-LoS, the study involved two stages. In the initial stage, various classifiers were employed including classical models such as Logistic Regression, ExtraTrees, Random Forest, KNN, Support Vector Classifier, as well as ensemble models like AdaBoost, GradientBoosting, XGBoost, and CatBoost. Random Forest yielded the highest validation accuracy at 0.94. In the subsequent stage, the Voting Classifier—an ensemble method—was applied, resulting in accuracy increasing to 0.96. Our method outperformed existing studies in terms of accuracy, including both neonatal-specific length of stay prediction studies and other general length of stay prediction research. While the length of stay estimation offers insights into the potential suitability of the incubators in the NICUs, which are not universally available in every city, for patient admission, it plays a pivotal role in delineating the treatment protocols of patients. Additionally, the research provides crucial information to the hospital management for planning such as beds, equipment, personnel, and costs. © The Author(s) 2024.
A Novel Approach for Image Compression Based on Multi-Level Image Thresholding Using Discrete Wavelet Transform and Cricket Algorithm
(Institute of Electrical and Electronics Engineers Inc., 2015) Canayaz, M.; Karci, A.
Applications of image compression is important in terms of time and resource management considering factors such as require more time to send according to the size of image over the network and large amount of space is high dimensional data for storing images. In this study, a new approach can be using at image compression process will be introduced. Firstly, image subjected to discrete wavelet transform for extracting feature. Then multi-level threshold values will be find with Shanon entropy in the obtained image. The maximum value of objective function will be obtained with the help of cricket algorithm at the threshold values finding step. This algorithm is a meta-heuristic algorithm that based on population. The threshold values that obtained through algorithm using to compressing the images will be provided. At the end of the study, the image compression ratio, the proposed approach running on a standard test image will be given. © 2015 IEEE.
A Novel Deep Learning-Based Approach for Prediction of Neonatal Respiratory Disorders From Chest X-Ray Images
(Elsevier B.V., 2023) Erdogan Yıldırım, A.; Canayaz, M.
In recent years, many diseases can be diagnosed in a short time with the use of deep learning models in the field of medicine. Most of the studies in this area focus on adult or pediatric patients. However, deep learning studies for the diagnosis of diseases in neonatal are not sufficient. Also, since it is known that respiratory disorders such as pneumonia have a large place among the causes of neonatal death, early and accurate diagnosis of respiratory diseases in neonates is crucial. For this reason, our study aims to detect the presence of respiratory disorders through the developed deep-learning approach using chest X-ray images of patients hospitalized in the Neonatal Intensive Care Unit. Accordingly, the enhanced version of C+EffxNet, the new hybrid deep learning model, is designed to predict respiratory disorders in neonates. In this version, the features selected by PCA are combined as 100, 200, and 300, then the binary classification process was carried out. In the study, the accuracy and kappa value were obtained as 0.965, and 0.904, respectively before feature merging, while these values were obtained as 0.977, and 0.935 after feature merging. This method, which was developed for the diagnosis of respiratory disorders in neonates, was also subsequently applied to a chest X-ray dataset that is frequently used in the literature for the diagnosis of pediatric pneumonia. For this data set, while the accuracy was 0.992, the kappa value was 0.982. The results obtained confirm the success of the proposed method for both datasets. © 2023 Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences
Removal of Thymol Blue From Aqueous Solution by Natural and Modified Bentonite: Comparative Analysis of Ann and Anfis Models for the Prediction of Removal Percentage
(University of Tehran, 2021) Koyuncu, H.; Aldemir, A.; Riza Kul, A.; Canayaz, M.
In this study natural bentonite (NB) and acid-thermal co-modified bentonite (MB) were utilized as adsorbents for the removal of Thymol Blue (TB) from aqueous solution. The batch adsorption experiments were conducted under different experimental conditions. The artificial neural network (ANN) and adaptive neuro fuzzy inference systems (ANFIS) were applied to estimate removal percentage (%) of TB. Mean squared error (MSE), root mean square error (RMSE) and coefficient of determination (R2) values were used to evaluate the results. In addition, the experimental data were fitted isotherm models (Langmuir, Freundlich and Temkin) and kinetic models (pseudo first order (PFO), pseudo second order (PSO) and intra-particle diffusion (IPD)). The adsorption of TB on both the NB and MB followed well the PSO kinetic model, and was best suited Langmuir isotherm model. When the temperature was increased from 298 K to 323 K for 20 mg/L of TB initial concentration, the removal percentage of TB onto the NB and MB increased from 74.91% to 84.07% and 81.19% to 93.12%, respectively. This results were confirmed by the positive ΔH° values indicated that the removal process was endothermic for both the NB and MB. The maximum adsorption capacity was found as 48.7805 mg/g and 117.6471 mg/g for the NB and MB, respectively (at 323 K). As a result, with high surface area and adsorption capacity, the MB is a great candidate for TB dye removal from wastewater, and the ANFIS model is better than the ANN model at estimating the removal percentage of the dye. © 2021 University of Tehran.
Training Artificial Neural Network With Chaotic Cricket Algorithm
(Institute of Electrical and Electronics Engineers Inc., 2018) Canayaz, M.; Ozdag, R.
Artificial neural networks are decision-making mechanisms inspired by the human nervous system. Many areas are widely used such as medicine, economics, machine learning. These nets generate an output value by passing input values through the layers. The input values are firstly multiplied by some weight values and sent to activation functions, thus, training is realized. The purpose of this training is to ensure that appropriate weight values are found. Meta-heuristic methods are widely used to find these weight values. In this study, Chaotic Map Cricket Algorithm was used to find the weight values. The weights obtained during the algorithm's operation were sent to the network and the weight matrix providing the minimum error was obtained. Since it is the first study on artificial neural network training with the Cricket Algorithm, the performance of the algorithm has been tried to be shown on the appropriate data sets in comparison with the Particle Swarm Optimization. © 2018 IEEE.