The linear regression prediction method is used to predict the future signal with the known signals in the past to compensate the delay caused by the filtering, and the AIC information norm method is used to determine the step of the self-regression prediction model, to provide the lat.

For a sparse multi-path MIMO-OFDM channel,a generalized Akaike information criterion(GAIC) based channel estimation applied approach is proposed.

Nearest neighbor clustering algorithm based on AIC criterion;

First, method for initialization of wavelet basis in wavelet frames is proposed; second, according to the data distribution, a modified method for rough selection of wavelet basis is given; and the last, an adaptive projection (algorithm) combined with AIC criterion is used to purify the wavelet basis, meanwhile finishing the parameters (identification.

This paper uses the AIC criterion proposed by Akaike(1973) to choose the common factor s number in factor analysis model,answers the question at the end of paper[1] and gives some discussion results, also analyzes and studies the data in a comparative study in measuring stress among university faculty in China and Japan.