We contend that corruption must be detected as soon as possible so that corrective and preventive measures may be taken. Thus, we develop an early warning system based on a neural network approach, specifically self-organizing maps, to predict public corruption based on economic and political...

This paper tests the hypothesis that stress tests are primarily a function of the fundamental financial condition and operating environment of individual banks, rather than alternative adverse economic and financial scenarios imposed by regulators. We develop a novel early warning system based...

We develop a model of neural networks to study the bankruptcy of U.S. banks. We provide a new model to predict bank defaults some time before the bankruptcy occurs, taking into account the specific features of the current financial crisis. Based on data from the Federal Deposit Insurance...