Univariate (auto) correlation functions, bivariate (cross) correlation functions, and temporal transfer functionsCorrelation function analysis is a way to study the temporal structure of the signals, in the time lag domain. The univariate analysis of a single signal leads to a study of its autocorrelation function. The bivariate analysis of two signals jointly leads to both auto-and cross-correlation functions.

No abstract

A quadric neural network (QNN) is designed for financial times series data prediction which is basically derived from Gabor's theory of communication and cybernetics. The QNN architecture is useful technique for prediction due to non-linear term. The simulation results are carried out on real financial time series data, which indicate better performance when taking minimum values in the input data set.

No abstract