For the two-component Gaussian mixture, it is hard to work out the maximum likelihood estimates (MLEs) and hence difficulty in the influence analysis.

• 摘要对于两高斯混合分布，很难求参数的极大似然估计，当然也不便于影响分析。