然而，我们知道的是，人们做出的决定基于对概率的判断，而此他们在头脑中不断计算和精确这个判断——这叫做“概率推理”。
What is known, however, is that people make decisions based on probabilities, which are constantly being calculated and refined in their heads—something called “probabilistic inference”.
最后，如此一系列的演绎推理将能处理整个句子和段落，从而生成一个精确的文本意义的陈述。
At the end, a series of deductions like these would handle entire sentences and paragraphs to yield an accurate text-meaning representation.
一般信度网的精确推理是一个NPC问题。
Exact inference on general belief networks is a NPC problem.
只有通过应用逻辑推理和精确、慎重的测量我们才意识到我们的进步。
It is through the use of logic and of precise, careful measurement that we become aware of our progress.
良好使用者，有能力运用英语，虽然在某些情况下有时会发生不精确， 不适当和误解。大致上掌握了复杂的英语，可以理解详细的论证推理。
Has operational command of the language though with occasional inaccuracies and misunderstandings in some situations. Generally handles complex language well and understands detailed reasoning.
在知识的推理策略中引入不精确推理机制，保证了推理机制的实现。
Meanwhile the indistinct inference mechanism is introduced in the inference tactics of knowledge so that the reasoning mechanism is assuredly realized.
本系统根据心理咨询领域中知识很难精确表述的特点，采用了模糊知识表示方法和加权模糊推理方法。
Because knowledge in the psychological field has the feature of accurate statements difficulty, this paper applied of fuzzy knowledge representation method and the weighted fuzzy reasoning method.
详细介绍了其模糊知识表示、模糊关系代数及不精确推理的实现策略。
Its fuzzy knowledge representation, fuzzy relational algebra and the inaccurate reasoning implementation policies are described in detail.
联合树算法以其容易理解，适用范围广等特点成为目前应用最多的精确推理算法。
Because Junction tree algorithm is easy to understand and its range of application is broad, it has been the most used widely among the exact inference algorithms.
本文对基于证据理论的k -NN分类方法进行了修正，得到了基于证据推理模型的k - NN分类方法，使分类结果更加精确。
By revising a K-NN classification method based on evidence theory, a new K-NN classification method based on evidence reasoning model is got, which made the classification result more accurate.
语言就是形式化方法中具有代表性的一种形式规格说明，它语言精确、无二义性，可以用于推理和求精。
Z language is a formal specification which is representative in formal method. Z language is accurate and without different meanings. It could be used to discursion and refinement.
它不需要建立所求解问题的精确数学模型，而是根据已知相关知识进行分析、推理和计算，最后得出事情的最终结果。
It does not need to establish an accurate model for the problem to be solved, but biased former knowledge to analyze, ratiocinate and calculate, and to form the last results.
本文论述了MYCIN不精确推理在理论上所存在的问题，并结合与地震预报有关的实践进行了讨论。
This paper deals with some defects of MYCIN inexact inference in theory, which are discussed with the practice of earthquake prediction.
理论分析和实验结果表明本算法推理严密，在精确性和稳定性都有显著的改善，是一种理想的特征点匹配算法。
Theoretical analysis and experimental results show that the improved algorithm has rigorous consequence and better accuracy and stability which is a perfect feature points matching algorithm.
通过面向故障的反向推理和深度推理方法，完成故障的精确定位；
The precise fault location performance is achieved by reverse inference and deep inference facing the fault models.
它不需要建立所求解问题的精确数学模型，而是模仿人脑的推理能力，可以使许多复杂问题简化。
It does not need to establish an accurate model for the problem solved, but it imitate the logical reasoning of person's brain, and can simplify many complicated problems.
专家在推理的精确性、诊断时间、信息搜索质量、提出诊断假设的数量等指标上与新手有显著差异。
There are significant differences between experts and novices in the accuracy and time of diagnosis, the quality of the searched information and the number of diagnostic hypotheses, etc.
基于模糊推理技术和基于可信度方法实现非精确推理模型的设计。
Realization the the imprecise reasoning model based on the Fuzzy reasoning technology and the Certain factors method.
它能精确地描述数字电路的时间特性，支持连续和离散的时间结构并能对时间信息进行推理。
Compared with propositional temporal logic, it can accurately describe the time feature of digital circuits and support discrete and continuous time structure.
在此基础上提出了基于多种概率的模糊因果图混合推理方法，以处理精确概率值、语言值和模糊概率值等多种概率。
In this paper, a mixed reasoning method of single-value FCD was proposed to deal with the reasoning of CD with multifold probabilities such as fuzzy Numbers, lingual values, accurate probabilities.
本文将专家系统中的MYCIN不精确推理方法推广应用到地震预报中。
This paper applies the MYCIN inexact inference method to the comprehensive earthquake prediction.
为改善周期精确级功耗分析的准确度和速度问题，使用多维特征参数建立贝叶斯推理的动态功耗模型。
To improve the accuracy and speed in cycle-accurate power estimation, this paper USES multiple dimensional coefficients to build a Bayesian inference dynamic power model.
提出一种用专家系统的不精确推理来得出评估结果的新方法，此方法可以应用于煤层的评估及矿井设计前的模拟。
This way can be used not only in the field of spontaneous combustion risk assessment in coal seam, but also in simulation before designing.
本文提出了一种模糊控制算法。利用传感器采集外部环境信息，在辨识获得精确模型的基础上，基于模糊推理，使机器人能够顺利完成火源追踪等任务。
In this paper, a fuzzy control algorithm was applied. With the sensors' messages, based on precise identification result and fuzzy inference, the robot can track fire successfully.
为了提高系统的鲁棒性，采用了基于可信度因子的不精确推理机制。
To improve system's robustness, uncertain reasoning based on certainty factor is introduced.
不精确推理和非单调推理是人工智能中两个重要的研究方面。
This paper analyzes the characteristics and limitations of approximate reasoning (based on probability) and nonmonotonic reasoning (based on truth maintenance system).
由于在雷达情报侦察中存在着很大的不确定性，因此在雷达辐射源的识别当中也存在着很大的不确定性，为此，本文选用不精确推理方式。
There are more uncertainties in the radar intelligence reception, so there are more uncertainties in the recognition of the radar radiation and the uncertainty reasoning is used in the paper.
逆向不精确推理用于实际工作场合，提供具有较高可信度的推理结果。
Converse precise inference is used in education and demonstraion, which can examine a…
提出了基于牵引计算和规则集推理的精确系统初始映像方法。
This paper proposes a new accurate simulation initialization method based on train performance calculation and rules-reasoning.
提出了基于牵引计算和规则集推理的精确系统初始映像方法。
This paper proposes a new accurate simulation initialization method based on train performance calculation and rules-reasoning.
应用推荐