Dincer H.Hacioglu U.Emir S.2024-03-132024-03-132014978331901125733190112439783319011240https://doi.org/10.1007/978-3-319-01125-7_8https://hdl.handle.net/20.500.12662/2991The main purpose of this study is to analyze the role of financial determinants on the bank profits in the Turkish Banking Sector. A comparative analysis has been conducted to predict the bank profits using Support Vector Regression (SVR) and Linear Regression (LR) models. The results illustrate that Net Interest Income After Specific Provisions/Total Operating Income, Non-Interest Income/Non-Interest Expense, Provision For Loan or Other Receivables Losses/Total Assets predictors have the most relative importance on SVR while Non-Interest Income/Non-Interest Expense, Provision For Loan or Other Receivables Losses/Total Assets predictors have it on LR. On the datasets containing these predictors, performances of SVR and LR models were compared based on Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) metrics. The findings present that SVR predicts the level of bank profits better than classical LR model based on both metrics. © 2014 Springer International Publishing Switzerland. All rights reserved.eninfo:eu-repo/semantics/closedAccessBankingFinancial determinantsLinear regressionProfitsSupport vector regressionBook Chapter10.1007/978-3-319-01125-7_82-s2.0-84930246633123N/A1099783319011257