Yazar "Atalan, Yasemin Ayaz" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    Consideration of Environmental, Economic, and Oil Factors for Unit-based Estimation of Consumed Electrical Energy with ML Algorithms: A Case Study of Çanakkale Region
    (Bursa Teknik Üniversitesi, 2025) Atalan, Yasemin Ayaz
    This study focuses on predicting electricity unit prices in the Çanakkale region by analyzing the effects of environmental, economic, and oil-related factors through machine learning (ML) algorithms. The research addresses the accurate prediction of energy costs amid fluctuating market dynamics by applying Random Forest (RF) and k-nearest neighbor (kNN) algorithms to monthly data from 2015 to 2024. The independent variables used in the models include exchange rate (USD/TRY), oil price (TL/liter), Producer Price Index (PPI), Consumer Price Index (CPI), and average temperature. The RF algorithm achieves superior predictive accuracy with an MSE of 0.013, RMSE of 0.112, MAE of 0.081, MAPE of 0.087, and an R² of 0.919, outperforming the kNN model across all metrics. The findings reveal that exchange rate and PPI have the most significant influence on electricity pricing. This study provides empirical evidence supporting the use of ML methods in energy price prediction and contributes to developing more accurate and robust forecasting tools for regional energy management and policy-making.
  • Küçük Resim Yok
    Öğe
    Integration of Machine Learning Algorithms and Discrete-Event Simulation for the Cost of Healthcare Resources
    (Mdpi, 2022) Atalan, Abdulkadir; Sahin, Hasan; Atalan, Yasemin Ayaz
    A healthcare resource allocation generally plays a vital role in the number of patients treated (p(nt)) and the patient waiting time (w(t)) in healthcare institutions. This study aimed to estimate p(nt) and w(t) as output variables by considering the number of healthcare resources employed and analyze the cost of health resources to the hospital depending on the cost coefficient (delta(i)) in an emergency department (ED). The integration of the discrete-event simulation (DES) model and machine learning (ML) algorithms, namely random forest (RF), gradient boosting (GB), and AdaBoost (AB), was used to calculate the estimation of the output variables depending on the delta(i) of resources cost. The AB algorithm performed best in almost all scenarios based on the results of the analysis. According to the AB algorithm based on the delta(0.0), delta(0.1), delta(0.2), and delta(0.3), the accuracy data were calculated as 0.9838, 0.9843, 0.9838, and 0.9846 for p(nt); 0.9514, 0.9517, 0.9514, and 0.9514 for w(t), respectively in the training stage. The GB algorithm had the best performance value, except for the results of the delta(0.2) (AB had a better accuracy at 0.8709 based on the value of delta(0.2) for p(nt)) in the test stage. According to the AB algorithm based on the delta(0.0), delta(0.1), delta(0.2), and delta(0.3), the accuracy data were calculated as 0.7956, 0.9298, 0.8288, and 0.7394 for p(nt); 0.8820, 0.8821, 0.8819, and 0.8818 for w(t) in the training phase, respectively. All scenarios created by the delta(i) coefficient should be preferred for ED since the income provided by the p(nt) value to the hospital was more than the cost of healthcare resources. On the contrary, the w(t) estimation results of ML algorithms based on the delta(i) coefficient differed. Although w(t) values in all ML algorithms with delta(0.0) and delta(0.1) coefficients reduced the cost of the hospital, w(t) values based on delta(0.2) and delta(0.3) increased the cost of the hospital.