一种基于集成学习的试题多知识点标注方法

个性化试题推荐、试题难度预测、学习者建模等教育数据挖掘任务需要使用到学生作答数据资源及试题知识点标注,现阶段的试题数据都是由人工标注知识点。因此,利用机器学习方法自动标注试题知识点是一项迫切的需求。针对海量试题资源情况下的试题知识点自动标注问题,本文提出了一种基于集成学习的试题多知识点标注方法。首先,形式化定义了试题知识点标注问题,并借助教材目录和领域知识构建知识点的知识图谱作为类别标签。其次,采用基于集成学习的方法训练多个支持向量机作为基分类器,筛选出表现优异的基分类器进行集成,构建出试题多知识点标注模型。最后,以某在线教育平台数据库中的高中数学试题为实验数据集,应用所提方法预测试题考察的知识点,取得了较好的效果。     

Powered partial least squares discriminant analysis

From the fundamental parts of PLS‐DA, Fisher's canonical discriminant analysis (FCDA) and Powered PLS (PPLS), we develop the concept of powered PLS for classification problems (PPLS‐DA). By taking advantage of a sequence of data reducing linear transformations (consistent with the computation of ordinary PLS‐DA components), PPLS‐DA computes each component from the transformed data by maximization of a parameterized Rayleigh quotient associated with FCDA. Models found by the powered PLS methodology can contribute to reveal the relevance of particular predictors and often requires fewer and simpler components than their ordinary PLS counterparts. From the possibility of imposing restrictions on the powers available for optimization we obtain an explorative approach to predictive modeling not available to the traditional PLS methods. Copyright © 2008 John Wiley & Sons, Ltd.     

Studying teachers’ mathematical argumentation in the context of refuting students’ invalid claims

This study investigates teachers’ argumentation aiming to convince students about the invalidity of their mathematical claims in the context of calculus. 18 secondary school mathematics teachers were given three hypothetical scenarios of a student's proof that included an invalid algebraic claim. The teachers were asked to identify possible mistakes and explain how they would refute the student's invalid claims. Two of them were also interviewed. The data were analysed in terms of the content and structure of argumentation and the types of counterexamples the teachers generated. The findings show that teachers used two main approaches to refute students’ invalid claims, the use of theory and the use of counterexamples. The role of these approaches in the argumentation process was analysed by Toulmin's model and three types of reasoning emerged that indicate the structure of argumentation in the case of refutation. Concerning the counterexamples, the study shows that few teachers use them in their argumentation and in general they underestimate their value as a proof method.     

K/n(G)系统可靠性评定的多源验前信息融合方法

在多源验前信息的情况下,以k/n(G)系统为例,讨论了如何对系统的可靠性指标进行估计的问题.在多个验前信息源给出了系统可靠性指标点估计的情况下,利用多层Bayes方法及经验Bayes方法对这些数据进行融合,并给出系统可靠性指标的Bayes点估计,同时对系统可靠度的置信区间也作了讨论.仿真算例表明这种处理方法是合理有效的.     

问卷设计知识管理中的范例推理

将范例推理技术应用于问卷设计知识重用过程中,提出了基于范例推理的问卷设计知识重用系统的应用框架,研究了问卷设计CBR系统中的一些关键技术:范例的表示、范例组织、范例检索算法、范例修正优化和重用、范例学习等.范例检索采用KNN法与遗传禁忌算法相结合的混合检索机制,通过建立GA适应度函数模型以及选择、禁忌交叉、禁忌变异算子等自动确定权重.我们组织和建立了初步的范例知识库,进行了相关实验.结果表明,融合遗传算法的范例检索方法可较大程度上缩短设计时间、提高问卷设计效率和质量,并有效地支持问卷设计的知识重用和知识管理创新.     

一种改进的图像快速去雾新方法

为提高雾霭天降质图像的清晰度和色彩保真度,基于暗原色先验提出了一种改进的图像快速去雾方法.针对原方法运算时间过长的问题,在保证去雾效果的前提下,利用快速双边滤波方法代替软件抠图修复透过率图,大幅度地降低了计算量和复杂度,处理时间仅为原方法的5％,满足了一般工程上的实时性要求.针对去雾处理后的图像亮度降低、颜色较实际场景偏暗的问题,提出了一种简单有效的图像对比度和亮度增强的方法,自适应地增强了图像的亮度,并对局部由于雾浓度过高而造成不清晰的区域,进行了对比度修正.结果表明,该方法能够快速有效地复原出雾天场景的对比度和真实色彩,在一定程度上保证了户外成像系统在恶劣天气下工作的稳定性和可靠性.     

Students’ ratings of teacher practices

In this paper, we explore a novel approach for assessing the impact of a professional development programme on classroom practice of in-service middle school mathematics teachers. The particular focus of this study is the assessment of the impact on teachers’ employment of strategies used in the classroom to foster the mathematical habits of mind and mathematical self-efficacy of their students. We describe the creation and testing of a student survey designed to assess teacher classroom practice based primarily on students’ ratings of teacher practices.     

Closure-based constraints in formal concept analysis

We present a method of imposing constraints in extracting formal concepts (equivalently, closed itemsets or fixpoints of Galois connections) from a binary relation. The constraints are represented by closure operators and their purpose is to mimic background knowledge a user may have of the data. The idea is to consider and extract only these itemsets that are compatible with the background knowledge. As a result, the method extracts less clusters, those that are interesting from the user point of view, in a shorter time. The method makes it also possible to extract minimal bases of attribute dependencies in the input data that are compatible with the background knowledge. We provide examples of several particular types of constraints including those that appeared in the literature in the past and present algorithms to compute the constrained formal concepts and attribute implications.     

Penalized Regressions: The Bridge versus the Lasso

Abstract

Bridge regression, a special family of penalized regressions of a penalty function Σ|β_j|^γ with γ ≤ 1, considered. A general approach to solve for the bridge estimator is developed. A new algorithm for the lasso (γ = 1) is obtained by studying the structure of the bridge estimators. The shrinkage parameter γ and the tuning parameter λ are selected via generalized cross-validation (GCV). Comparison between the bridge model (γ ≤ 1) and several other shrinkage models, namely the ordinary least squares regression (λ = 0), the lasso (γ = 1) and ridge regression (γ = 2), is made through a simulation study. It is shown that the bridge regression performs well compared to the lasso and ridge regression. These methods are demonstrated through an analysis of a prostate cancer data. Some computational advantages and limitations are discussed.