Browsing by Author "Arslan, Ilker"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Conference Object
    Determination of Alzheimer's Disease Levels by Ordinal Logistic Regression and Artificial Learning Algorithms
    (Ieee, 2024) Bulut, Nurgül; Çakar, Tuna; Arslan, Ilker; Akinci, Zeynep Karaoglu; Oner, Kevser Setenay
    This study compares artificial learning algorithms and logistic regression models in determining different levels of Alzheimer's disease (AD). The research uses demographic, genetic, and neurocognitive inventory results obtained from the National Alzheimer's Coordination Center (NACC) database, along with brain volume/thickness measurements derived from MRI scanners. Deep Neural Networks, Ordinal Logistic Regression, Random Forest, Gaussian Naive Bayes, XGBoost, and LightGBM models were employed to determine the 4 different ordinal levels of AD. Although there were similarities between the accuracy rate, F1 score, AUC value, and sensitivity, specificity, and precision performance measures of each class, the highest classification rate was achieved by the Random Forest model where the oversampling was not applied. (F1 score: 0.86; accuracy: 0.86 and AUC: 0.95). The outputs of the model with the best performance were explained with the SHAP (SHapley Additive exPlanations) method. These findings indicate that non-invasive markers and artificial learning models can be used effectively in early diagnosis and decision support systems to predict different levels of Alzheimer's disease.
  • Thumbnail Image
    Conference Object
    Feature Enrichment Via Similar Trajectories for Xgboost Based Time Series Forecasting
    (Ieee, 2024) Yilmaz, Elif; Islak, Umit; Çakar, Tuna; Arslan, Ilker
    In this study, new time series forecasting models are developed based on XGBoost, and the similar trajectories method (ST), which can be interpreted as a regression based on nearest neighbors. Both the similar trajectories method and XGBoost model are known to have successful applications in traffic flow prediction. In our case, the focus is on similar trajectories used in the former method, and features based on these trajectories are used in the training of XGBoost. The success of the proposed models is confirmed through metrics such as the mean absolute error. Also, statistical tests are performed among the compared benchmark models. The study is concluded with discussions and questions about how these models can be further developed.