Evaluating Bayesian networks’ precision for detecting students’ learning styles

Students are characterized by different learning styles, focusing on different types of information and processing this information in different ways. One of the desirable characteristics of a Web-based education system is that all the students can learn despite their different learning styles. To achieve this goal we have to detect how students learn: reflecting or acting; steadily or in fits and starts; intuitively or sensitively. In this work, we evaluate Bayesian networks at detecting the learning style of a student in a Web-based education system. The Bayesian network models different aspects of a student behavior while he/she works with this system. Then, it infers his/her learning styles according to the modeled behaviors. The proposed Bayesian model was evaluated in the context of an Artificial Intelligence Web-based course. The results obtained are promising as regards the detection of students’ learning styles. Different levels of precision were found for the different dimensions or aspects of a learning style.     

AH-questionnaire: An adaptive hierarchical questionnaire for learning styles

One of the main concerns when providing learning style adaptation in Adaptive Educational Hypermedia Systems is the number of questions the students have to answer. Most of the times, adaptive material available will discriminate among a few categories for each learning style dimension. Consequently, it is only needed to take into account the general tendency of the student and not the specific score obtained in each dimension. In this context, we present AH-questionnaire, a new approach to minimize the number of questions needed to classify student Learning Styles. Based on the Felder-Silverman’s Learning Style Model, it aims at classifying students into categories in spite of providing precise scores. The results obtained in a case study with 330 students are very promising. It was possible to predict students’ learning style preference with high accuracy and only a few questions.     

Learning styles and formative assessment strategy: enhancing student achievement in Web-based learning

Abstract The purpose of this research was to investigate the effects of formative assessment and learning style on student achievement in a Web-based learning environment. A quasi-experimental research design was used. Participants were 455 seventh grade students from 12 classes of six junior high schools. A Web-based course, named BioCAL, combining three different formative assessment strategies was developed. The formative assessment strategies included Formative Assessment Module of the Web-Based Assessment and Test Analysis system (FAM-WATA) (with six Web-based formative assessment strategies), Normal Module of Web-Based Assessment and Test Analysis system (N-WATA) (only with partial Web-based formative assessment strategy) and Paper and Pencil Test (PPT) (without Web-based formative assessment strategy). Subjects were tested using Kolb's Learning Style Inventory, and assigned randomly by class into three groups. Each group took Web-based courses using one of the formative assessment strategies. Pre- and post-achievement testing was carried out. A one-way ANCOVA analysis showed that both learning style and formative assessment strategy are significant factors affecting student achievement in a Web-based learning environment. However, there is no interaction between these two factors. A post hoc comparison showed that performances of the FAM-WATA group are higher than the N-WATA and PPT groups. Learners with a 'Diverger' learning style performed best followed by, 'Assimilator', 'Accommodator', and 'Converger', respectively. Finally, FAM-WATA group students are satisfied with six strategies of the FAM-WATA.     

Ontology based approach to Bayesian student model design

Probabilistic student model based on Bayesian network enables making conclusions about the state of student’s knowledge and further learning and teaching process depends on these conclusions. To implement the Bayesian network into a student model, it is necessary to determine “a priori” probability of the root nodes, as well as, the conditional probabilities of all other nodes. In our approach, we enable non-empirical mathematical determination of conditional probabilities, while “a priory” probabilities are empirically determined based on the knowledge test results. The concepts that are believed to have been learned or not learned represent the evidence. Based on the evidence, it is concluded which concepts need to be re-learned, and which not. The study described in this paper has examined 15 ontologically based Bayesian student models. In each model, special attention has been devoted to defining “a priori” probabilities, conditional probabilities and the way the evidences are set in order to test the successfulness of student knowledge prediction. Finally, the obtained results are analyzed and the guidelines for ontology based Bayesian student model design are presented.     

A Bayesian Diagnostic Algorithm for Student Modeling and its Evaluation

In this paper, we present a new approach to diagnosis in student modeling based on the use of Bayesian Networks and Computer Adaptive Tests. A new integrated Bayesian student model is defined and then combined with an Adaptive Testing algorithm. The structural model defined has the advantage that it measures students' abilities at different levels of granularity, allows substantial simplifications when specifying the parameters (conditional probabilities) needed to construct the Bayesian Network that describes the student model, and supports the Adaptive Diagnosis algorithm. The validity of the approach has been tested intensively by using simulated students. The results obtained show that the Bayesian student model has excellent performance in terms of accuracy, and that the introduction of adaptive question selection methods improves its behavior both in terms of accuracy and efficiency.     

基于预测关系的贝叶斯网络学习算法

在介绍有代表性的贝叶斯网络结构学习算法基础上,给出了变量之间预测能力的概念及估计方法,并证明了预测能力就是预测正确率,在此基础上建立了基于变量之间预测关系的贝叶斯网络结构学习方法,并使用模拟数据进行了对比实验,实验结果显示该算法能够有效地进行贝叶斯网络结构学习。     

The max-min hill-climbing Bayesian network structure learning algorithm

We present a new algorithm for Bayesian network structure learning, called Max-Min Hill-Climbing (MMHC). The algorithm combines ideas from local learning, constraint-based, and search-and-score techniques in a principled and effective way. It first reconstructs the skeleton of a Bayesian network and then performs a Bayesian-scoring greedy hill-climbing search to orient the edges. In our extensive empirical evaluation MMHC outperforms on average and in terms of various metrics several prototypical and state-of-the-art algorithms, namely the PC, Sparse Candidate, Three Phase Dependency Analysis, Optimal Reinsertion, Greedy Equivalence Search, and Greedy Search. These are the first empirical results simultaneously comparing most of the major Bayesian network algorithms against each other. MMHC offers certain theoretical advantages, specifically over the Sparse Candidate algorithm, corroborated by our experiments. MMHC and detailed results of our study are publicly available at http://www.dsl-lab.org/supplements/mmhc_paper/mmhc_index.html. Editor: Andrew W. Moore     

Internet project-based learning environment: the effects of thinking styles on learning transfer

The purpose of this study, in an environment of Internet project‐based learning, is to undertake research on the effects of thinking styles on learning transfer. In this study, we establish an environment that incorporates project‐based learning and Internet. Within this environment, we divide our sample of elementary school students into four groups: Executive Group, Legislative Group, Judicial Group, and Mixed Group. Taking the learning of ‘Natural Science’ as an example, we investigate the effects of different thinking styles on learning transfer. The results of this study are:

• (a) significant differences between the near transfer of the Executive Group and the Legislative Group,
• (b) no significant differences in far transfer are found among groups of different thinking styles,
• (c) the near transfer of the Mixed Group is superior to that of the Legislative Group and the Judicial Group, and
• (d) the far transfer of the Mixed Group is superior to that of the Legislative Group.

     

混合式贝叶斯网络结构增量学习算法研究

提出一种混合式贝叶斯网络结构增量学习算法.首先提出多项式时间的限制性学习技术,为每个变量建立候选父节点集合;然后,依据候选父节点集合,利用搜索技术对当前网络进行增量学习.该算法的复杂度显著低于目前最优的贝叶斯网络增量学习算法.理论与实验均表明,所处理的问题越复杂,该算法在计算复杂度方面的优势越明显.     

A hybrid algorithm for Bayesian network structure learning with application to multi-label learning

We present a novel hybrid algorithm for Bayesian network structure learning, called H2PC. It first reconstructs the skeleton of a Bayesian network and then performs a Bayesian-scoring greedy hill-climbing search to orient the edges. The algorithm is based on divide-and-conquer constraint-based subroutines to learn the local structure around a target variable. We conduct two series of experimental comparisons of H2PC against Max–Min Hill-Climbing (MMHC), which is currently the most powerful state-of-the-art algorithm for Bayesian network structure learning. First, we use eight well-known Bayesian network benchmarks with various data sizes to assess the quality of the learned structure returned by the algorithms. Our extensive experiments show that H2PC outperforms MMHC in terms of goodness of fit to new data and quality of the network structure with respect to the true dependence structure of the data. Second, we investigate H2PC’s ability to solve the multi-label learning problem. We provide theoretical results to characterize and identify graphically the so-called minimal label powersets that appear as irreducible factors in the joint distribution under the faithfulness condition. The multi-label learning problem is then decomposed into a series of multi-class classification problems, where each multi-class variable encodes a label powerset. H2PC is shown to compare favorably to MMHC in terms of global classification accuracy over ten multi-label data sets covering different application domains. Overall, our experiments support the conclusions that local structural learning with H2PC in the form of local neighborhood induction is a theoretically well-motivated and empirically effective learning framework that is well suited to multi-label learning. The source code (in R) of H2PC as well as all data sets used for the empirical tests are publicly available.     

基于因果影响独立模型的贝叶斯网络参数学习

基于因果影响独立模型及其中形成的特定上下文独立关系,提出一种适于样本学习的贝叶斯网络参数学习算法。该算法在对局部概率模型降维分解的基础上,通过单父节点条件下的子节点概率分布来合成局部结构的条件概率分布,参数定义复杂度较低且能较好地处理稀疏结构样本集。实验结果表明,该算法与标准最大似然估计算法相比,能充分利用样本信息,具有较好的学习精度。     

Being Bayesian About Network Structure. A Bayesian Approach to Structure Discovery in Bayesian Networks

In many multivariate domains, we are interested in analyzing the dependency structure of the underlying distribution, e.g., whether two variables are in direct interaction. We can represent dependency structures using Bayesian network models. To analyze a given data set, Bayesian model selection attempts to find the most likely (MAP) model, and uses its structure to answer these questions. However, when the amount of available data is modest, there might be many models that have non-negligible posterior. Thus, we want compute the Bayesian posterior of a feature, i.e., the total posterior probability of all models that contain it. In this paper, we propose a new approach for this task. We first show how to efficiently compute a sum over the exponential number of networks that are consistent with a fixed order over network variables. This allows us to compute, for a given order, both the marginal probability of the data and the posterior of a feature. We then use this result as the basis for an algorithm that approximates the Bayesian posterior of a feature. Our approach uses a Markov Chain Monte Carlo (MCMC) method, but over orders rather than over network structures. The space of orders is smaller and more regular than the space of structures, and has much a smoother posterior landscape. We present empirical results on synthetic and real-life datasets that compare our approach to full model averaging (when possible), to MCMC over network structures, and to a non-Bayesian bootstrap approach.     

基于双层贝叶斯网增强学习机制的网络认知算法

为提高网络认知的准确度,采用双层贝叶斯网络模型对网络参数进行层次化描述;采用强化学习推理算法对模型的条件概率表进行分级和学习,删除冗余信息,更准确地反映网络参数间的依赖关系,保证网络认知算法的准确度。经仿真分析,证明算法能够更好地描述网络参数的依赖信息,具有较高的推理准确度。     

Bayesian networks for student model engineering

Bayesian networks are graphical modeling tools that have been proven very powerful in a variety of application contexts. The purpose of this paper is to provide education practitioners with the background and examples needed to understand Bayesian networks and use them to design and implement student models. The student model is the key component of any adaptive tutoring system, as it stores all the information about the student (for example, knowledge, interest, learning styles, etc.) so the tutoring system can use this information to provide personalized instruction. Basic and advanced concepts and techniques are introduced and applied in the context of typical student modeling problems. A repertoire of models of varying complexity is discussed. To illustrate the proposed methodology a Bayesian Student Model for the Simplex algorithm is developed.     

A novel Bayesian deep learning methodology for enhanced foetal cardiac signal mining

Cardiac defects are amongst the most common birth defects. Cardiac diagnosis is indispensably imperative in the foetal stage as it might help provide an opportunity to plan and manage the baby during Antepartum and Intrapartum stages, when the baby is born. It is from the Antepartum stage where the foetal electrocardiogram (fECG) signal can actually be detected. At present, monitoring the foetus is completely focused on the heart rate. Currently fECG analysis is used in the clinical domain to analyse heart rate and the allied variations. Analysis using the morphology of the fECG is generally not undertaken for cardiac-anomaly populations. The ultimate reason for this scenario is due to unavailability in technology to yield trustworthy fECG measurements with desired quality required by Physicians. A novel hybrid methodology called BDL (Bayesian Deep Learning) methodology is proposed. The BDL includes a Bayesian filter and a deep learning (DL) Artificial Intelligent neural network for maternal electrocardiogram (mECG) elimination and non-linear artefacts removal to yield high quality non-invasive fECG signal. The outcomes of the research by the proposed BDL system proved valuable and provided high quality fECG signal for efficient foetal diagnosis.     

What type of learning style leads to online participation in the mixed-mode e-learning environment? A study of software usage instruction

Learning style is traditionally assumed to be a predictor of learning performance, yet few studies have identified the mediating and moderating effects between the two. This study extends previous research by proposing and testing a model that examines the mediating processes in the relationship between learning style and e-learning performance and the moderating effects of prior knowledge. The results show that the sensory/intuitive dimension of learning style predicts learning performance indirectly through the mediation of online participation. However, other types of learning styles do not affect online participation. Sensory students demonstrate a higher level and intuitive students a lower level of online participation. Prior knowledge plays an important role as a moderator between online participation and learning performance. This study was conducted in the context of software usage instruction using empirical data from 219 undergraduate students.     

贝叶斯网络在学生模型建模中的应用分析

将课程教学资源融合到学生模型构建中,描述了包括领域知识拓扑结构的建立、条件概率表学习算法的推理的详细过程,最终得到了学生模型中关于章节知识项的贝叶斯网络结构图,并通过一个实验系统对个性化教学系统中学生模型建构的整个框架的可行性进行了验证.     

贝叶斯网络结构学习综述

贝叶斯网络是一种有效的不确定性知识表达和推理工具,在数据挖掘等领域得到了较好的应用,而结构学习是其重要研究内容之一.经过二十多年的发展,已经出现了一些比较成熟的贝叶斯网络结构学习算法,对迄今为止的贝叶斯网络结构学习方法进行了综述.现阶段获得的用于结构学习的观测数据都比较复杂,这些数据分为完备数据和不完备数据两种类型.针对完备数据,分别从基于依赖统计分析的方法、基于评分搜索的方法和混合搜索方法三个方面对已有的算法进行分析.对于不完备数据,给出了数据不完备情况下网络结构的学习框架.在此基础上归纳总结了贝叶斯网络结构学习各个方向的研究进展,给出了贝叶斯网络结构学习未来可能的研究方向.