The more unexpected the knowledge mined, the more valuable . the notable characteristic that bayes method has is that the hypothesis can be reflected by the result . if the previous knowledge is known little, or even unknown, bayes method has its peculiar merit 而贝叶斯(bayes)方法的一个显著特点,就是它可以通过看结果来了解假设,也就是说,当对先验知识知之甚少,或者毫不知情的情况下,贝叶斯方法具有其它方法不可比拟的长处。
The more unexpected the knowledge mined, the more valuable . the notable characteristic that bayes method has is that the hypothesis can be reflected by the result . if the previous knowledge is known little, or even unknown, bayes method has its peculiar merit 而贝叶斯(bayes)方法的一个显著特点,就是它可以通过看结果来了解假设,也就是说,当对先验知识知之甚少,或者毫不知情的情况下,贝叶斯方法具有其它方法不可比拟的长处。
Based on the above background, at first, thesis analyzes the characteristics and risk in ship investment . secondly, it presents two kinds of utility function about risk-averse decision-makers . combining with bayes method, author set up ship investment risk evaluation model 基于以上背景,本文首先分析了船舶投资的特点及投资项目所面临的四类风险,结合效用理论给出两种风险厌恶型投资者的效用函数,运用贝叶斯分析方法建立一种新的船舶投资风险模型。
In this paper, an integral scheme of 16 position error calibration and autonomous alignment for three axis platform is given . it may calibrate 33 errors in all . first, determine parameters with least square estimate, then bayes method, ridge regression estimation were discussed separately 本文设计了一个十六位置误差标定方案,可以分离出总计33项误差,首先用最小二乘估计方法进行参数辨识,而后,分别研究了基于bayes方法的误差系数辨识,基于岭估计的误差系数辨识。
On practical occasions, censored test is used, small scale sample test theory and extensive sample strategy are applied, together with point estimation, estimation of lower limit value for reliability in certain confidence and bayes method based on ? ii ? abstract prior reliability information of the old system are adopted to estimate and veflf ~ ? the new system ? reliability 根据现场具体情况,采用截尾试验方法,应用小子样试验理论,采用扩大样本量策略,利用点估计、在一定置信度下系统可靠度下限估计和以原有系统可靠性信息为先验信息的bayes法对系统的可靠性进行估计和验证。
First of all, least square method and modified maximum likelihood estimation of zero-failure data for the exponential distribution, the weibull distribution, the normal distribution and the logarithm normal distribution are given, among them, the bayes method in the estimation to the failure probability is introduced 首先,分别采用最小二乘法和修正极大似然估计法对指数分布、weibull分布、正态及对数正态分布的无失效数据处理进行了研究,其中,在对失效概率的估计中引入了bayes方法。
In this paper, according to the characteristic and essence of data mining and bayes method, bayes theory are applied to the methods of data mining such as clustering, classification, association rule and abnormity analysis . some algorithms are brought forward and verified and discussed 本文主要针对数据挖掘的特点和本质,充分利用贝叶斯方法的特点,将贝叶斯理论及其思想方法融入到聚类、分类、关联规则挖掘以及偏差分析和异常检测等数据挖掘各方法中,提出将贝叶斯方法应用于数据挖掘的算法,并对提出的算法进行了验证和讨论。
Secondly, in respect bayes method is difficult in extracting features and in establishing statistical models, and the capability of no-linear approach is preferably for bp neural netw ork, the method ( bpbayes method ) to combine neural network with subjective bayesian was introduced . in this method, bp neural network was used to replace the experts of relevant domains for applying subjective bayes method 其次,考虑到bayes方法在特征建模上具有较大的难度,而bp神经网络具有较强的非线性逼近能力,为此我们将bp神经网络的输出代替领域专家的知识经验,并用bp神经网络与主观bayes方法结合起来的bpbayes方法对目标进行识别。
Secondly, in respect bayes method is difficult in extracting features and in establishing statistical models, and the capability of no-linear approach is preferably for bp neural netw ork, the method ( bpbayes method ) to combine neural network with subjective bayesian was introduced . in this method, bp neural network was used to replace the experts of relevant domains for applying subjective bayes method 其次,考虑到bayes方法在特征建模上具有较大的难度,而bp神经网络具有较强的非线性逼近能力,为此我们将bp神经网络的输出代替领域专家的知识经验,并用bp神经网络与主观bayes方法结合起来的bpbayes方法对目标进行识别。