Extensions of LDA by PCA mixture model and class-wise features

Linear discriminant analysis (LDA) is a data discrimination technique that seeks transformation to maximize the ratio of the between-class scatter and the within-class scatter. While it has been successfully applied to several applications, it has two limitations, both concerning the underfitting problem. First, it fails to discriminate data with complex distributions since all data in each class are assumed to be distributed in the Gaussian manner. Second, it can lose class-wise information, since it produces only one transformation over the entire range of classes. We propose three extensions of LDA to overcome the above problems. The first extension overcomes the first problem by modelling the within-class scatter using a PCA mixture model that can represent more complex distribution. The second extension overcomes the second problem by taking different transformation for each class in order to provide class-wise features. The third extension combines these two modifications by representing each class in terms of the PCA mixture model and taking different transformation for each mixture component. It is shown that all our proposed extensions of LDA outperform LDA concerning classification errors for synthetic data classification, hand-written digit recognition, and alphabet recognition.     

A motivational model of rural students' intentions to persist in, versus drop out of, high school.

Using self-determination theory, the authors tested a motivational model to explain the conditions under which rural students formulate their intentions to persist in, versus drop out of, high school. The model argues that motivational variables underlie students' intentions to drop out and that students' motivation can be either supported in the classroom by autonomy-supportive teachers or frustrated by controlling teachers. LISREL analyses of questionnaire data from 483 rural high school students showed that the provision of autonomy support within classrooms predicted students' self-determined motivation and perceived competence. These motivational resources, in turn, predicted students' intentions to persist, versus drop out, and they did so even after controlling for the effect of achievement. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

The value of certification: an empirical assessment of the perceptions of end-users of local area networks

Abstract. Certification is one criterion that employers seek when hiring network professionals. Many employers assume a certified network professional is better able to manage networked resources than is a non-certified counterpart. Therefore, employers often offer a monetary premium to attract certified network professionals. By measuring end-user perceptions, the current study empirically assesses the popular industry assumption that certified network professionals are more adept managers of network resources than those individuals who are not certified network professionals. Analyses of a data derived from 299 end-users of 11 separate local area networks indicate no statistically significant difference exists between the capabilities of certified network professionals and non-certified network professionals.     

在线磨辊辊面粗糙度的实验研究

影响轧辊表面粗糙度的因素很多，为此对影响辊面粗糙度的因素进行了系统分析和实验研究，得出了轧辊和砂轮的速度、砂轮粒度、砂轮直径及磨削液等主要因素对轧辊表面粗糙度影响的一般规律。利用回归分析的方法建立了轧辊表面粗糙度的数学模型，并将计算值与实测值进行了比较，发现两者基本一致。此项研究为适时控制冷轧带钢在线磨辊辊面粗糙度提供了计算模型。     

武钢 10000 m3/h制氧机主冷泄漏故障处理及应急保产方式

氧气生产单位保用户用气一旦出现较为紧张的局面时，如何正确处理保安全生产运行、保用户生产、安排抢修时间三者之间的关系，值得关注与研究。     

转炉合金最小成本控制模型

介绍了基于合金最小成本的转炉炼钢钢水成分调整模型工应用于宝钢炼钢在线生产控制的情况。模型给出了在满足目标成分等工艺要求下，成本最小的合金投入组合及合金投入量；计算了合金加入钢包后引起钢水温度、重量等的变化，为降低生产成本（合金投入），提高出钢成分控制精度提供了有效的途径。     

热带钢连轧机工作辊温度场和热凸度预报模型

采用一种具有较高精度的样条有限条法对轧制过程中工作辊的温度场和热凸度进行计算和预报。结果证明，该模型理论计算值与现场实测值吻合较好，可以较准确地计算和预报热带钢连轧机工作辊的热凸度。     

Localization of Function Within the Dorsal Hippocampus: The Role of the CA3 Subregion in Paired-Associate Learning.

Computational models and electrophysiological data suggest that the CA3 subregion of the hippocampus supports the formation of arbitrary associations; however, no behavioral studies have been conducted to test this hypothesis. Rats with neurotoxin-induced lesions of dorsal dentate gyrus (DG), CA3, or CA1 were tested on object-place and odor-place paired-associate tasks to test whether the mechanism that supports paired-associate learning is localized to the CA3 subregion of the dorsal hippocampus or whether all hippocampal subregions contribute to paired-associate learning. The data indicate that rats with DG or CA1 lesions learned the tasks as well as controls; however, CA3-lesioned rats were impaired in learning the tasks. Thus, the CA3 subregion of the dorsal hippocampus contains a mechanism to support paired-associate learning. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Microscale and mesoscale discrete models for dynamic fracture of structures built of brittle material

In this work we present the discrete models for dynamic fracture of structures built of brittle materials. The models construction is based on Voronoi cell representation of the heterogeneous structure, with the beam lattice network used to model the cohesive and compressive forces between the neighboring cells. Each lattice component is a geometrically exact shear deformable beam which can describe large rigid body motion and the most salient fracture mechanisms. The latter can be represented through the corresponding form of the beam constitutive equations, which are derived either at microscale with random distribution of material properties or at a mesoscale with average deterministic values. The proposed models are also placed within the framework of dynamics, where special attention is paid to constructing the lattice network mass matrix as well as the corresponding time-stepping schemes. Numerical simulations of compression and bending tests is given to illustrate the models performance.     

Decreasing miscibility with greater heterogeneity in ternary polymer blend: poly(cyclohexyl methacrylate), poly(α‐methyl styrene) and poly(4‐methyl styrene)

A ternary blend system comprising poly(cyclohexyl methacrylate) (PCHMA), poly(α‐methyl styrene) (PαMS) and poly(4‐methyl styrene) (P4MS) was investigated by thermal analysis, optical and scanning electron microscopy. Ternary phase behaviour was compared with the behaviour for the three constituent binary pairs. This study showed that the ternary blends of PCHMA/PαMS/P4MS in most compositions were miscible, with an apparent glass transition temperature (T_g) and distinct cloud‐point transitions, which were located at lower temperatures than their binary counterparts. However, in a closed‐loop range of compositions roughly near the centre of the triangular phase diagram, some ternary blends displayed phase separation with heterogeneity domains of about 1 µm. Therefore, it is properly concluded that ternary PCHMA/PαMS/P4M is partially miscible with a small closed‐loop immisciblity range, even though all the constituent binary pairs are fully miscible. Thermodynamic backgrounds leading to decreased miscibility and greater heterogeneity in a ternary polymer system in comparison with the binary counterparts are discussed. © 2003 Society of Chemical Industry