Abstract: In the light of the irregular probability density functions needed for Bayes Parameter Estimation, a new method is provided for estimating the related probability density functions by using Parzen Windows Estimation. So, Bayes Parameter Estimator can be obtained on choosing different error function as the cost function. By choosing a group of resistance measurements as the datasets, the probability density functions are derived by Parzen Windows Estimation methods and Minimum Mean Absolute Error （MMAE） estimator of a resistance dataset is obtained directly with Cross-Validations.

Key words: Parzen Window estimation, Bayes estimation, Cross-Validations

摘要： 针对贝叶斯估计中所需的非规则概率密度函数, 提出用Parzen窗算法估计相关概率密度, 从而求解不同损失函数下的贝叶斯参数估计器. 实例分析中, 选择一组电阻测量值作为样本, 利用Parzen窗法计算出相应的概率密度函数, 最后用交叉验证法得出了该样本的最小绝对值误差参数估计器.

关键词: Parzen窗法, 贝叶斯估计, 交叉验证法.