Transformations  in
Business & Economics

THE HIERARCHICAL DETERMINATION OF CUSTOMER SATISFACTION WITH BANKING SERVICES USING AN ARTIFICIAL NEURAL NETWORK
Margareta Ilie, Andreea-Daniela Moraru, Silvia Ghita-Mitrescu

ABSTRACT. During the last two decades, especially after the set-off of the financial crisis in 2007, the competition within the banking industry has become fiercer than ever, yielding a constant preoccupation for customer satisfaction. This paper aims at approaching customer satisfaction in the Romanian banking system from a new perspective, using artificial intelligence techniques. To this end, the research focuses on building and training an Artificial Neural Network (ANN) in order to simulate bank customer satisfaction and find the criteria that are most important for training the ANN. The authors used a feed forward ANN and set the error between the real output and the simulated output at less than 10%. The ANN simulated values based on the data recorded from a survey with more than 600 participants who assessed different satisfaction criteria related to banking services. During the training process, the ANN determined the most important criteria for bank customer satisfaction.

KEYWORDS: customer satisfaction, banking services, Artificial Neural Network, Romania.

JEL classification: M31, G21, C45.