Adaptive Test for White Noise
针对传统白噪声检验对某些备择假设不敏感的问题，提出一种自适应检验方法，先利用数据有限傅里叶变换的实部选择合适频率，再检验这些频率上虚部的显著性，适用于任意平稳备择假设。
Since conventional tests for white noise are completely insensitive to certain alternatives, it is difficult in a particular application to decide which test to use if one has no a priori information about the kind of departure from whiteness. Therefore, an alternative test is proposed which may be applied in case of arbitrary stationary alternatives. This test procedure is adaptive in that first the real part of the finite Fourier transform of the data is used to choose appropriate frequencies and then the significance of the imaginary part at these frequencies is tested.