IPC is a key question of distributed system.

Meanwhile, the classification way of Maxent is adjusted for the numerical classification of audio classification and sentence segmentation.

同时，针对音频分类和句子切分的数值分类方式，对Maxent分类器的分类方式进行调整。

互联网摘选

The spatial-temporal variation of species diversity has significant influence on the MaxEnt method to estimate species richness.

结果表明，物种多样性的时空异质性对于估计物种丰富度的最大熵方法有比较明显的影响。

互联网摘选

At last, MaxEnt and GARP models were compared.

最后比较研究了MaxEnt模型和GARP模型。

互联网摘选

MaxEnt was not always effective and was inclined to give an upper limit of species richness estimated in a single area.

最大熵方法并不总是有效的，它倾向于给出一个地区物种丰富度的上限值。

互联网摘选

MaxEnt should be used with caution because of some restrictions, including sampling area at the anchor scale, sample plot number, historical disturbance factors, evenness of species distribution and heterogeneity of sample unit.

同时，在使用最大熵方法的时候必须注意一些影响其估计精度的限制性条件，包括锚定格局上的取样面积、调查的样方数目、历史干扰因素、物种分布的均匀程度及取样的异质性。

互联网摘选

The first innovations of thesis is that through in-depth analysis on advantage of collaborative training algorithm co-training, in minimizing the amount of label data and guaranteeing the classification accuracy, the co-training algorithm based on maximum entropy ( Maxent) is used for audio classification.

本文的一个创新点是，通过深入分析协同训练算法co-training在极小化标注数据量并保证分类精度方面的优势，采用基于最大熵分类的co-training算法进行音频分类。

互联网摘选

Prediction of potential geographic distribution areas of the maize downy mildew in China by using MAXENT

利用MAXENT预测玉米霜霉病在中国的适生区

互联网摘选

Random Iterated Algorithm for HMMS Based on EM and Selected Application in Biology Sequence Promoter Recognition

基于EM的隐马氏过程随机迭代算法及其在生物序列启动子识别中的应用

互联网摘选

Each individual in evolutionary computation represents a set of HMMs, while the fitness value of each individual represents the maximum mutual information.

各模型集用个体表示，个体的适应值采用模型的最大互信息。

互联网摘选

A calculation method of the similar degree of two HMMs is introduced based on entropy concept.

介绍了基于信息熵概念定义的两个隐马氏模型的相似度的一个计算方法。

互联网摘选

And this has led to the local optimality of the models. This paper proposed a hybrid maximum mutual information/ evolutionary computation ( MMI/ EC) architecture to be embedded in the training of HMMs.

文中将最大互信息（MMI）和演化计算（EC）相结合，引入到隐马尔柯夫模型（HMM）的训练中去。

互联网摘选

Hidden Markov models ( HMMs) are used as representations for the cluster centers, and batch map training algorithm is applied in clustering procedure.

隐马尔可夫模型被用于表达各个聚类，批处理自组织特征被用于符号序列的聚类过程。

互联网摘选

This paper presents the results of training HMMs using a minimum error criterion ( ME) instead of conventional maximum likelihood.

本文研究了使用最小差错准则训练隐含马尔科夫模型（HMMS）而不是传统最大似然准则。

互联网摘选

Bank Bill Automation Recognition Based on HMMs Method

基于HMM方法的银行票据自动识别

互联网摘选

To study multiple sequence alignment and profile HMMs, one essential algorithm of HMMs and how to build an HMMs are discussed.

探讨了多重序列比对和序列谱隐马尔可夫模型，讨论了隐马尔可夫模型的基本算法以及如何建立HMMs。

互联网摘选

Multi-aspect Radar Target Recognition Method Based on Scattering Centers and HMMs Classifiers

基于目标散射中心和HMM分类的多视角雷达目标识别方法

互联网摘选

But the problem is that the two parameter estimation algorithms for HMMs, the Viterbi algorithm and Baum Welch algorithm, are only guaranteed to find a local optimum.

然而，目前的HMM参数估计算法Viterbi算法和Baum-Welch算法，都只能找到局部最优比对，无法找到全局最优比对。

互联网摘选

A double k-mean clustering algorithm for modeling and clustering the gene sequence data is proposed by using the hidden Markov models ( HMMs).

本文提出了一种基于隐马尔可夫模型的二次k-均值聚类算法并实现了对基因序列数据的建模与聚类。

互联网摘选

First, a set of HMMs are trained with historical process data to construct a database. Each HMM corresponds to a fault.

首先，使用过程历史数据训练得到各种过程故障工况对应的HMM；

互联网摘选

An Improving calculation Method of the HMMs Similar Degree

一个计算隐马氏模型相似度的改进方法

互联网摘选