Yazar "Calgici, E." seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    AI-Based Classification of Normal and Aggressive Behaviors using EMG Signals
    (Institute of Electrical and Electronics Engineers Inc., 2023) Karakoc, M.; Calgici, E.; Kandaz, D.; Ucar, M.K.
    Behavior analysis using Electromyogram (EMG) signals is an essential step in understanding aggressive behaviors, the physiological behavior of the neuromuscular system, and its applications in other disciplines. Therefore, this study aimed to develop a model for detecting behaviors using EMG signal analysis. This model utilized a dataset consisting of EMG signals from eight channels obtained during a series of activities on four subjects. Accordingly, a classification model was developed to detect specific features of behaviors and differentiate between normal and aggressive behaviors by analyzing the EMG signals. The model was developed by incorporating signal and statistical features and applying different machine-learning techniques. A total of 44 models were evaluated for their performance. The Support Vector Machine (SVM) classification model developed using all extracted features achieved an accuracy rate of approximately % 91 in behavior classification. The obtained results demonstrate the potential of EMG signals as a tool for behavior detection and classification. © 2023 IEEE.
  • Küçük Resim Yok
    Öğe
    Hybrid AI-Based Chronic Kidney Disease Risk Prediction
    (Institute of Electrical and Electronics Engineers Inc., 2023) Yordan, H.H.; Karakoc, M.; Calgici, E.; Kandaz, D.; Ucar, M.K.
    There are various models and traditional methods for predicting the risk of chronic kidney disease(CKD), which can enhance treatment effectiveness, slow down disease progression, and reduce the risk of complications. However, these methods have limitations. In this study, a hybrid risk prediction model based on artificial intelligence is proposed to optimize the treatment process using data from individuals diagnosed with CKD. The dataset consists of 29 attributes, including medical laboratory results and patient history. Data sets created by utilizing these attributes in specific proportions were tested in classification models. By combining K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Ensemble Bagged Tree (EBT) machine learning algorithms, a Hybrid Machine Learning (HML) model was developed. The hybrid prediction model can accurately predict kidney disease risk with % 100 accuracy. Therefore, a supportive model for clinicians in the diagnosis and treatment process has been achieved. © 2023 IEEE.