Advancing the Backtrack Optimization Technique to Obtain Forecasts of Potential Crisis Periods

Abstract

Financial crisis is an unfortunate reality that overshadows any financial system regardless its profitability and the level it functions. The appearance of crises across financial markets, especially during the 1990s that the internationalized markets adopted a rather approachable character, imposed severe costs in financial and social systems. With this paper is proposed the generation of a future interval of time that is vulnerable to enclose the burst of a financial crisis. A time series consisted of approximations of the local Lipschitz constant is examined and in the proposed forecasting approach this constant holds the crisis indicator role. Further the application of two different optimization techniques over the Lipschitz-made time series results to the generation of a future period of time; this interval is likely to envelop the burst of a forthcoming crisis. The usage of a future interval of time empowers the predicting ability of the methodology by providing warning signs priory to the actual crisis burst. To this direction, the obtained results offer strong evidence that the method may be characterized as an Early Warning System (EWS) for financial crisis prediction

An Intelligent Method Based on State Space Search for Automatic Test Case Generation

“…Backtrack [14] is an optimum seeking method which searches forward according to the selection conditions to achieve a goal. If at a certain step of the search, it is found that the goal turns out to be unachievable, then a step backward is taken.…”

An Intelligent Method for Test Data Generation Based on Optimized Interval Arithmetic

“…A CSP is generally solved by search strategies, among which backtracking algorithms [27] are widely used. In this paper, the process of exploring the solution space is represented by state space search [28] in artificial intelligence.…”