分类者

Results: Fifty-two of 66 patients with unilateral fractures were, after 9–13 years, rated as excellent/good according to the Green and O'Brien score as modified by Cooney et al.

结果：66例单侧桡骨无故骨折患者，有52例在伤后9－13年GOBC评分为优良，AO骨折分类与结果无关，相当多的患者仍遗留骨折移位：背侧成角（小于60岁者，平均13°，大于60岁者，平均18°），与伤时相比桡骨存在短缩（小于60岁者，平均2mm，大于60岁者，平均3mm），5例患者复位后仍存关节面缺损（1－2mm），但是仅有1例患者后来发展成了轻度的骨性关节炎。

After investigation, the personality characteristics of our countrymen are classified into eight groups: Religious Skeptics (18.6%), Indifferentism (7.9%), Thrill-lovers (18.8%), Self-pursuers (6.8%), Makers (11.4%), Attention-getters (16.0%), Retreaters (12.3%) and Ordinary people (8.3%).

此八项分类分别为「宗教怀疑者」(18.6%)、「冷淡者」(7.9%)、「追求刺激者」(18.8%)、「自我追求者」(6.8%)、「实做者」(11.4%)、「聚焦者」(16.0%)、「退缩者」(12.3%)及「平实者」(8.3%)。

In order to decrease the error rate of speech recognition, speaker adaptation techniques are often used to adjust speaker-dependent acoustic models. MLLR (Maximum Likelihood Linear Regression) and MAP are two of the most popular techniques in recent years. MLLR uses the technique of regression trees. It calculates the transform matrix for each leaf node of the tree. This makes it possible to use fewer sentences to decrease the error rate of HMM-based speech recognition. However, while we examined the recognition result, we found that although the overall error rate decreased, but the error rate of certain confusable phones was higher. In order to solve this problem, we propose the use MPE (Minimum Phone Error Discriminative Training) to solve this problem.

在语音辨识技术的实际应用上，语者调适技术常被用做调整语者无关之声学模型，使其对於特定语者之辨识率有所提升，常见的语者调适技术有最大化相似度线性回归（Maximum Likelihood Linear Regression），其精神在於透过对语音辨识模型相近的mixture做群聚，再对各个群聚做调整，以期能达到利用少量语料，提升辨识率的目的，但其缺点为，在发音上相近的模型，其mixture原本就极为近似，若再分类为同一群聚做调整，容易因为使用者的发音习惯，而使模型偏向ㄓ或ㄗ，因此虽然整体的辨识率获得提升，但却造成混淆音的错误率上升。

Results Of the 52 specimens, 38 were located in the ciliary body (73.1%), and 14 lesions involved the iris (26.9%). According to the results of histogenetic classification, the tumors of ciliary sensory epithelium were ranked the first place (20 cases, 38.5%) while malignant melanoma, the second one (14 cases, 26.9%) followed by melanocytoma (9 cases, 17.3%) and granulomatous lesions (5 cases, 9.6%).

结果 52例组织标本按解剖部位分类：位于睫状体病变38例（73.1%），原发于睫状体累及虹膜组织14例（26.9%）；按组织来源分类：占第1位者为睫状体神经上皮来源的肿瘤20例（38.5%），第2位者为恶性黑色素瘤14例（26.9%），第3和第4位者分别为黑色素细胞瘤9例（17.3%）和肉芽肿性炎性病变5例（9.6%）。

The key finding indicates that of the participants who received categorized WOM, those who received information displayed in a specific forum are likely to be more subjective on their purchase decisions than those who received information displayed in a integrated forum. Participants tend to be more secure, unconfused, and less likely to anticipate regret for their purchase decision when receiving with categorized information.

最后研究发现，当受测者所浏览的口碑资讯有受到分类，并以「个别讨论区」的方式呈现时，其「决策主观状态」会较无口碑资讯分类、以「综合讨论区」的方式呈现之组别要来的好，亦即有资讯分类，较能使受测者对决策感到确定、不困惑且预期未来发生后悔的可能性较低。

By applying Talmy's semantic theory of causation, this paper divides the Ba-construction into five categories in terms of its semantic structure, namely, the causing-event Ba-construction, the agent Ba-construction, the author Ba-construction, the instrument Ba-construction, and the undergoer Ba-construction, the major type of which is subdivided into some smaller groups according by their syntactic structures.

本文以Talmy的致使义理论为框架，从致使义的角度重新考察&把&字句的语义，将&把&字句按其语义类型分成致使事件&把&字句，施动者&把&字句，起始者&把&字句，工具&把&字句以及遭遇者&把&字句，并根据其语法结构对其中主要类型做了更细致的分类。

Pulse of been extracted in different scales of energy feature is classified by fuzzy cluster analysis and come into being standard pulse model bank, which is used to help unsorted pulse find the closest pulse in standard pulse model bank by fuzzy diagnosis.

故采用模糊聚类分析的方法对已提取不同尺度能量特征的几种脉象进行分类并形成标准脉象模型库，供下一步模糊模式识别对未分类的脉象在标准脉象模型库中找出最接近者。

Widely employed in paper handling equipment, copiers and sorters, these clutches need to engage and disengage quickly and precisely.

广泛被用于纸张处理设备，复印机和分类者，这些离合器需要担并且迅速而确切停止交战。

Most divergences between classifications of the same substance should be removed over time either through co-operation between notifiers and registrants or by EU harmonized classifications for substances that are category 1, 2, and 3 CMRs, respiratory sensitizers, and other substances if justified on a case-by-case basis.

大多数对同种物质的分类的分歧须通过通报者或注册者之间的合作或由欧盟的统一分类即分I、II、III类CMRs物质，呼吸道感光剂及其他经个案一一证实的物质。

The first is the introduction of narratology and characteristics of TV program of in-depth news reporting as narrative works.

第一章主要为叙事学理论介绍和电视新闻深度报道节目作为叙事作品的特性；第二章主要是对电视新闻深度报道中叙事者的分类；第三章是关于电视新闻深度报道中叙事者的结构方式和叙事者的功能。

Objective: To study the effect of polycythemia on blood oxygen saturation.

裴蕾目的：观察RBC剧增而引起的高粘血症对血氧饱和度的影响。

Based on SIMPLER algorithm in the curvilinear body-fitted coordinates, the calculations were performed for Pr=0.7, Re=10~1000 on non-orthogonal non-staggered grids which are generated by elliptic equation systems.

采用曲线坐标系下压力与速度耦合的SIMPLER算法，数值研究了周期性渐扩渐缩波纹通道内脉动流动与换热情况，流动Re数的范围为10~1000，Pr数为0.7。