文章介绍了数理统计中实数或向量参数的正函数的后验平均、方差和预测密度的近似计算。
This paper introduces the approximate calculation of real number's or of vector parameter positive function's posterior mean, variance and forecasting density in mathematical statistics.
频谱参数的预测公式由各基频下的条件概率密度函数导出,频谱参数的修改通过正弦模型实现。
The spectral parameters were predicted using the conditional probabilistic distribution function based on the pitch and were modified synchronizingly with the pitch using the sinusoidal model.
当状态过程具有马氏性时,还分别给出滤波、预测以及内插的条件分布密度函数的递推格式。
When the state process has Markov property, the recursive formulae for the conditional distribution density functions of filtering, prediction. and interpolation, are given respectively.
应用推荐