The economic crises in the world have also affected the banking sector and caused an increase in bank failures. Therefore, predicting bank failures as earlier as possible has become more important to take the necessary precautions in advance. This book aims at developing early-warning models to predict bank failures up to three years prior to failure and examines the case of Turkey. The models are developed using two different data mining techniques: logistic regression analysis and neural networks. The financial ratios derived from the financial statements of the banks are used to construct the models. The results show that capital adequacy, asset quality, liquidity position, profitability, and income-expenditure structure of a bank are the indicators of its likelihood of failure at a posterior time. Besides the bank failure prediction models, this book also gives a review of data mining techniques and mainly focuses on the factor analysis, logistic regression analysis, and neural networks. The book is intended to help the bank supervisors, bank balance sheet analysts, and investors, as well as the readers interested in the banking sector and also data mining techniques.