历史查询   



 为了更好的帮助您理解掌握查询词或其译词在地道英语中的实际用法，我们为您准备了出自英文原文的大量英语例句，供您参考。 
However, Bayesian networks model the entire distribution of each conditional probability and could consume significant computational resources.


This is the focus of the paper: we propose here to train the structured Bayesian Network model using a crossentropy training criterion.


To compare the alternative approaches in practice, each of the predictive distributions discussed is instantiated in the Bayesian network model family case.


A Bayesian Network model of intrusion to be monitored by this agent is provided as an argument to this agent on startup.


A dynamic Bayesian network model for autonomous 3d reconstruction from a single indoor image Erick Delage Honglak Lee Andrew Y.


A fitness function evaluates the performance of each individual, a Bayesian network model in this case, to measure how close it is to the solution.

 更多 


 As a knowledge representation framework and a kind of probability inference engine, Bayesian networks are widely used in applications for reasoning and decision making with inherent uncertainty. Since the exact algorithms of probability inference in Bayesian networks is NPhard, as the topology of the network becomes more dense, the runtime complexity of probabilistic inference increases dramatically and realtime decision making eventually becomes prohibitive, so many approximate algorithms based on simulation... As a knowledge representation framework and a kind of probability inference engine, Bayesian networks are widely used in applications for reasoning and decision making with inherent uncertainty. Since the exact algorithms of probability inference in Bayesian networks is NPhard, as the topology of the network becomes more dense, the runtime complexity of probabilistic inference increases dramatically and realtime decision making eventually becomes prohibitive, so many approximate algorithms based on simulation or model simplification are proposed. The method discussed in this paper is based on the model simplification of arc removal. In this method, a subset of arcs are selected and removed, which simplifies the network structure and we obtain an approximate network, then any probability inference algorithm can be applied to this approximate network to get a solution within the error bound we predefined. By using the KullbackLeibler information divergence as the measure of the difference between two probability distributions, this paper discusses the multiple arc removal problem in the gen eral case and presents the optimal parameters for the approximate network. Final ly, a heuristic algorithm is provided which searches a set of arcs to be removed under the upper bound on the probability error allowed.  弧的删除是一种对 Bayes网络模型进行近似的方法 .文中以 Kullback L eibler偏差作为近似网络和原网络概率分布误差的测度 ,给出了近似网络在此测度意义下的最优参数 .同时 ,也给出了通过对原网络删除多条弧进行近似的启发式算法 ,当给定一个误差上界时 ,可以使用此算法寻找满足误差要求的近似网络  A learning approach is proposed to solve the problems of conditional probability assignation in large scale Bayesian network. Firstly, a new hierarchical Bayesian Network model is defined based on class hierarchical structure, which is used to represent large scale Bayesian network.Then, the train data set is changed from a single table to a database composed of some database tables. And each database table corresponds to a Bayesian network block. Based on that, a formula of conditional probability is developed.... A learning approach is proposed to solve the problems of conditional probability assignation in large scale Bayesian network. Firstly, a new hierarchical Bayesian Network model is defined based on class hierarchical structure, which is used to represent large scale Bayesian network.Then, the train data set is changed from a single table to a database composed of some database tables. And each database table corresponds to a Bayesian network block. Based on that, a formula of conditional probability is developed. And each conditional probabilistic table of Bayesian network block can be calculated from the database tables respectively. Properly adjust the attribute number in each database table can assure the validity of this learning approach. Finally, experiments in automatic detection and location of texts in images show the feasibility of this hierarchical Bayesian network and learning approach.  针对大规模Bayes网络的条件概率赋值问题,提出一种学习方法.首先使用类层次结构定义一种新的层次Bayes网络模型,用于表示大规模Bayes网络.然后将训练数据集由单个数据表的形式转化成多表数据库,其中每个数据库表对应1个Bayes网络模块.在此基础上导出条件概率计算公式,从每个数据库表中算出相应的Bayes网络模块的条件概率表,由此实现对整个层次Bayes网络的概率赋值.可通过适当增加数据库表的数目来控制每个表中属性的个数,保证计算的可行性.最后将本层次Bayes网络及计算公式用于解决图像中文本的自动检测与定位问题,实验结果表明了它们的有效性.  A learning approach is proposed to solve the problems of conditional probability assignation in large scale Bayesian networks.Firstly, a new hierarchical Bayesian Network model is defined based on class hierarchical structure,which is used to represent large scale Bayesian networks.Then,the train data set is changed from a single table to a database composed of some database tables.And each database table corresponds to a Bayesian network block.Based on that,a formula of conditional probability is developed.And... A learning approach is proposed to solve the problems of conditional probability assignation in large scale Bayesian networks.Firstly, a new hierarchical Bayesian Network model is defined based on class hierarchical structure,which is used to represent large scale Bayesian networks.Then,the train data set is changed from a single table to a database composed of some database tables.And each database table corresponds to a Bayesian network block.Based on that,a formula of conditional probability is developed.And each conditional probabilistic table of Bayesian network block can be calculated from the database tables respectively.Proper adjustment of the attribute number in each database table can assure the validity of this learning approach.Experiments in automatic detection and location of texts in images show the feasibility of this hierarchical Bayesian network and learning approach.  针对大规模Bayes网络的条件概率赋值问题,提出一种学习方法.首先使用类层次结构定义一种新的基于层次的Bayes网络模型,用于表示大规模Bayes网络.然后将训练数据集由单个数据表的形式转化成多表数据库,其中每个数据库表对应一个Bayes网络模块.在此基础上导出条件概率计算公式,从每个数据库表中算出相应的Bayes网络模块的条件概率表,由此实现对整个层次Bayes网络的概率赋值.通过适当增加数据库表的数目来控制每个表中属性的个数,保证计算的可行性.将层次Bayes网络及计算公式用于解决图像中文本的自动检测与定位问题,实验结果表明了它们的有效性.   << 更多相关文摘 
相关查询  



