Abstract
The banking industry is regarded as the backbone of a country’s modern economy. Globalization and technological innovation have produced a highly competitive market in the banking and financial industry. Banks’ performance is heavily dependent on the accuracy of managerial judgments. The objective of this research is to predict bank performance using several types of artificial intelligence techniques. Prediction models are based on artificial intelligence techniques such as Artificial Neural Networks and MultiLayer Perceptron (ANN-MLP), Radial Basis Function (RBF) network, Random Forest (RF) technique and multiple linear regression (MLR). The data used in developing models includes 405 samples collected from 45 banks in Vietnam during the period 2002-2022. Predicted outputs are total loans and total deposits. Experimental results and model evaluation criteria indicate that the prediction model based on RF technique provides the highest accuracy. Specifically, the values of RMSE; MAPE; MAE; R; Theil’s U obtained by RF respectively are 6.2132×107;1.2763; 2.8180×107; 0.9651; 0.1498 (for total loans prediction) and 6.8408×107; 2.6092; 3.0893×107; 0.9741; 0.1469 (for total deposit prediction).
Keywords: Bank performance, Artificial Intelligence, Neural Network, Random Forest, RBF, Multiple Linear Regression, Vietnam.
JEL classification: E5, G21, G24.
Để đọc toàn văn vui lòng nhấn vào xem chi tiết.
Ban biên tập Tạp chí Kinh tế & Quản trị Kinh doanh
Phòng 514, Nhà điều hành, trường Đại học Kinh tế & Quản trị Kinh doanh
Địa chỉ: Phường Tân Thịnh, thành phố Thái Nguyên
Email: tapchikt-qtkd@tueba.edu.vn; Điện thoại: 0208.3903373
