后验概率

In this algorithm, the direction-of-arrival is assumed to be a discrete random variable with a priori probability density function defined on a set of candidate points. Whether or not the secondary sample is required is based on the a posteriori probability distribution of a set of candidate point"s, which can be calculated from the array received signals. And then, the resulting beamformer is a weighted sum of the beamformers pointed at the latest set of point"s, which are combined according to the value of the a posteriori probability for each pointing direction.

在该算法中，期望信号的波达方向被看作是一个定义在若干候选波达方向上的离散随机变量，由阵列接收信号计算出各候选波达方向的后验概率，并将后验概率的分布作为是否需要进行二次采样的判断标准；随后，对各波达方向上的波束成形权值线性加权，便可得到自适应波束成形器，权系数为各候选波达方向的后验概率值。

In this paper, a speckle reduction method for SAR images based on Dual Tree Complex Wavelet Transform is proposed. It first presumes the probability distribution function for complex wavelet coefficients of the original nature image, and then applies MAP theory to estimate coefficients from the noisy wavelet coefficients. Filtered image can be derived after inversed DTCWT.

该方法首先对无噪声污染图像的复数小波系数的统计概率分布进行建模；然后利用此先验概率模型，采用最大后验概率方法从含噪小波系数中估计出无噪声污染的小波系数：最后经重构得到滤波后的图像。

From statistical point of view, this is a Bayesian one in which a posterior probability density surface is calculated for points on a grid in the parameter space.

从统计观点上讲，这是一种在参数空间对每个栅格点计算后验概率密度的Bayesian统计，采用平滑函数对不规则的后验概率密度曲面进行平滑化，从而在置信的95尹。

This paper presents a novel multiple object segmentation algorithm based on a Bayesian framework. According to the characteristic of the intra-frame and inter-frame information, a representation of Maximization of the A posteriori Probability of spatio-temporal label field is proposed. So a minimization of energy function is obtained.

摘要该文提出了一种基于贝叶斯框架的时空标记场最大后验概率的多视频对象分割算法，根据视频序列帧间和帧内信息的不同特点，建立基于多个对象分割标记场的最大后验概率公式，并导出其最小能量函数，通过求解最小能量使其分割标记的后验概率达到最大。

A weighted posterior probability method is presented to calculate the probability outputs of support vector machinesfor multi class cases.

中文摘要：为解决传统支持向量机方法不提供概率输出的问题，在支持向量机多类分类问题输出概率建模中，提出了加权后验概率建模方法。

ALSAB adopts Maximum A Posteriori provability estimation and Expectation-Maximization algorithm to estimate the weight parameters of LSA,and ALSAB employs incremental learning to decrease accumulative effect caused in continuous update with considering that the probability of uncommon words decreases in continuous update.

ALSAB采用最大后验概率估计与期望值最大算法对概率LSA模型参数进行有效的估计，在充分考虑多次更新中不常用字词概率参数降低问题的前提下，采用增量学习方法降低多次更新产生的累积效应。

Secondly, the properties of output sequences of stop-and-go clocked combiners are discussed. It is proved that the output sequences are stationary and obey strong law of large numbers and central limit theorem. Using the probabilistic model properly built, we analyze the rate of coincidence between the output sequences of two kinds of generators with basic operations and corresponding LFSRs sequences. All the computational formulae are derived. A recursive algorithm for efficiently computing the poster probability of partial input strings and corresponding joint probability is introduced. Moreover, divide-and-conquer attacks on this kind of generators which reconstruct the initial states of the input stop-and-go generators individually are proposed.

其次，分析了一般钟控停走组合生成器输出序列的概率分布性质，证明了生成器的输出序列是严平稳的，且服从强大数定律和中心极限定理；考察了钟控加法型组合生成器和钟控乘法型组合生成器的输出序列和相应的常规钟控下线性移位寄存器序列之间的符合率，给出了符合率的具体计算公式；给出了组合器输出序列段和输入序列段之间的联合分布以及部分输入序列段的后验概率的有效递归算法，进一步提出了利用后验概率对钟控组合生成器进行分别征服相关攻击的方法。

The computation of posterior probability considers the evidence E certain without dealing with uncertain evidence,in terms of this,the posterior probability is split the difference according to the probability of evidence appearing in Bayesian network and its extension network of Credal network,and is expressed by Credal set so as to provide the base of decision-making for decision-maker.

任何变量的后验概率计算都假定证据变量E是确定的，无法处理不确定证据下的后验概率。针对这个问题，提出在贝叶斯网络及其扩展的Credal网络中应用证据出现的概率对后验概率进行折中处理，并用不确定的Credal集表达其结果，从一定程度上为决策人员提供有意义的决策依据。

The computation of posterior probability considers the evidence E certain without dealing with uncertain evidence,in terms of this,the posterior probability is split the difference according to the probability of evidence appearing in Bayesian network and its extension network of Credal network,and is expressed by Credal set so as to provide the base of decision-making for decision-maker.

中文摘要：任何变量的后验概率计算都假定证据变量E是确定的，无法处理不确定证据下的后验概率。针对这个问题，提出在贝叶斯网络及其扩展的Credal网络中应用证据出现的概率对后验概率进行折中处理，并用不确定的Credal集表达其结果，从一定程度上为决策人员提供有意义的决策依据。

Monte Carlo is a method that approximately solves mathematic or physical problems by statistical sampling theory. When comes to Bayesian classification, it firstly gets the conditional probability distribution of the unlabelled classes based on the known prior probability. Then, it uses some kind of sampler to get the stochastic data that satisfy the distribution as noted just before one by one. At last, it can obtain the posterior probability distibution of each unlabelled classes by analysing these stochastic data.

蒙特卡罗是一种采用统计抽样理论近似求解数学或物理问题的方法，它在用于解决贝叶斯分类时，首先根据已知的先验概率获得各个类标号未知类的条件概率分布，然后利用某种抽样器，分别得到满足这些条件分布的随机数据，最后统计这些随机数据，就可以得到各个类标号未知类的后验概率分布。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。