A new reverse reduce-error ensemble pruning algorithm

Although greedy algorithms possess high efficiency, they often receive suboptimal solutions of the ensemble pruning problem, since their exploration areas are limited in large extent. And another marked defect of almost all the currently existing ensemble pruning algorithms, including greedy ones, consists in: they simply abandon all of the classifiers which fail in the competition of ensemble selection, causing a considerable waste of useful resources and information. Inspired by these observations, an interesting greedy Reverse Reduce-Error (RRE) pruning algorithm incorporated with the operation of subtraction is proposed in this work. The RRE algorithm makes the best of the defeated candidate networks in a way that, the Worst Single Model (WSM) is chosen, and then, its votes are subtracted from the votes made by those selected components within the pruned ensemble. The reason is because, for most cases, the WSM might make mistakes in its estimation for the test samples. And, different from the classical RE, the near-optimal solution is produced based on the pruned error of all the available sequential subensembles. Besides, the backfitting step of RE algorithm is replaced with the selection step of a WSM in RRE. Moreover, the problem of ties might be solved more naturally with RRE. Finally, soft voting approach is employed in the testing to RRE algorithm. The performances of RE and RRE algorithms, and two baseline methods, i.e., the method which selects the Best Single Model (BSM) in the initial ensemble, and the method which retains all member networks of the initial ensemble (ALL), are evaluated on seven benchmark classification tasks under different initial ensemble setups. The results of the empirical investigation show the superiority of RRE over the other three ensemble pruning algorithms.     

Ensemble cluster pruning via convex-concave programming

Ensemble learning is the process of aggregating the decisions of different learners/models. Fundamentally, the performance of the ensemble relies on the degree of accuracy in individual learner predictions and the degree of diversity among the learners. The trade-off between accuracy and diversity within the ensemble needs to be optimized to provide the best grouping of learners as it relates to their performance. In this optimization theory article, we propose a novel ensemble selection algorithm which, focusing specifically on clustering problems, selects the optimal subset of the ensemble that has both accurate and diverse models. Those ensemble selection algorithms work for a given number of the best learners within the subset prior to their selection. The cardinality of a subset of the ensemble changes the prediction accuracy. The proposed algorithm in this study determines both the number of best learners and also the best ones. We compared our prediction results to recent ensemble clustering selection algorithms by the number of cardinalities and best predictions, finding better and approximated results to the optimum solutions.     

基于区域重构的树状骨架快速去毛刺方法

区域骨架去毛刺是骨架提取与应用中的重要问题。常见的去毛刺途径之一,是用基于区域重构的骨架显著性指标对骨架进行阈值化处理,但存在算法参数难以直观设置、去毛刺效果不易控制、运行速度较慢等问题。针对上述问题,提出了一种逐次剪除骨架分枝的去毛刺方法,以突出部分骨架长度为显著性指标,每次剪除显著性最低的一个分枝,直至剩余分枝达到给定数量;为提高算法速度,采用了游程森林结构加速区域重构操作,提出了重构触发策略来减少重构次数。在实际图像集上的实验结果表明,提出的方法的正确骨架分枝的召回率较对比算法高13%,准确率高近3%;采用重构触发策略的算法运行时间平均为未采用该策略算法的约56%。实验结果表明了所提方法的有效性。     

Scale independent surface characterisation: Geography meets precision surface metrology

The recently published standard ISO 25178-2 distinguishes between field parameters and feature parameters for surface texture characterisation, whereby the main difference between these two types is due to the fact that the parameters belonging to the first group are deduced from all points of a scale-limited surface, while the parameters belonging to the second group are deduced from only a subset of predefined topological surface features. As specified in ISO 25178-2, two prerequisites are indispensable for the determination of the feature parameters, viz., an adequate data structure for surface characterisation and a suitable formal method for surface generalisation, i.e. for the successive elimination of the less important surface features. Within ISO 25178-2 change trees are proposed for describing the surface topography, while Wolf pruning is suggested for surface simplification (cf. also ISO 16610-85). Apart from the techniques specified in ISO 25178-2 and ISO 16610-85, the present paper describes a second geometrical–topological approach for the characterisation and generalisation of surfaces that has its origin within the geosciences and is based on weighted surface networks and w-contractions. In addition, it is revealed how the two approaches, both of which have their foundations in graph theory, are interrelated to each other and how, from a historical point of view, the GIScience approach forms the basis of the one applied within surface metrology. Finally, some applications within precision engineering are described.     

Use of tree pruning wastes for manufacturing of wood reinforced cement composites

Tree pruning wastes from six woody species, namely Acacia salicina, Conocarpus erectus, Ficus altissima, Leucaena glauca, Pithecellobium dulce and Tamarix aphylla, were used to manufacture high-quality wood reinforced cement composites (WRCCs). Hydrations experiments were conducted to screen the compatibility of the selected tree pruning wastes with cement. Additionally, various particle pretreatments and chemical additives were applied to enhance the compatibility of wood with cement. The best treatment for each species was selected and used to manufacture the WRCCs. The panels were produced under specific manufacturing variables and the mechanical properties and dimensional stability of the panels were determined. The results indicated that both board density and wood/cement (W/C) ratio had significant effects on the properties of WRCCs. With few exceptions, a W/C ratio of 1/2 and either 1200 kg m⁻³ or 1300 kg m⁻³ produced the optimal strength properties. The tree pruning wastes are suitable for use as raw materials in the manufacturing of WRCCs after pre-treatment of the wood particles with either cold or hot water and with addition 3% of CaCl₂, Al₂(SO₄)₃ or MgCl₂. Therefore, these wastes could be used as an alternative wood source for WRCCs.     

Sensitivity Analysis for Decision Boundaries

A novel approach is presented to visualize and analyze decision boundaries for feedforward neural networks. First order sensitivity analysis of the neural network output function with respect to input perturbations is used to visualize the position of decision boundaries over input space. Similarly, sensitivity analysis of each hidden unit activation function reveals which boundary is implemented by which hidden unit. The paper shows how these sensitivity analysis models can be used to better understand the data being modelled, and to visually identify irrelevant input and hidden units.     

决策树算法的优化研究

针对决策树C4.5/5.0分类算法及改进的算法在创建决策树时训练误差率和校验误差率相对较高的缺点,提出一些改进策略,即利用属性相关性进行属性约简与度量以达到解决属性集合中的冗余属性,采用一定置信度值进行决策树的修剪,采用优化的Chi2算法更合理更准确地对连续属性进行离散化,基于改进策略设计并实现一个分类器,将改进的算法应用于Breast-cancer实例,实验结果证明改进的算法生成的决策树具有较高的分类正确率。     

一种面向任务的对地观测卫星Agent团队构建方法

随着航天科技的飞速发展,逐渐出现了由多种异构卫星组成的卫星集群。相比于传统的卫星系统,卫星集群具有规模大、平台多、载荷异构的特点,传统的卫星任务规划方法难以适用。针对卫星集群任务规划中的关键问题——面向任务的卫星Agent团队构建问题,建立了数学模型,提出了基于分支限界的精确搜索算法,并对其时间复杂度进行了分析。针对精确算法时间复杂度较高的缺点,引入了启发式剪枝机制,并按照任务集合排序策略的不同设计了3种启发式卫星团队构建算法。最后,通过多组实验分析了卫星团队构建精确搜索算法与启发式剪枝搜索算法的性能,验证了我们提出算法的有效性和实用性。     

基于聚类的两段式孤立点检测算法

现有的大多数孤立点检测算法都需要预先设定孤立点个数,并且还缺乏对不均匀数据集的检测能力。针对以上问题,提出了基于聚类的两段式孤立点检测算法,该算法首先用DBSCAN聚类算法产生可疑孤立点集合,然后利用剪枝策略对数据集进行剪枝,并用基于改进距离的孤立点检测算法产生最可能孤立点排序集合,最终由两个集合的交集确定孤立点集合。该算法不必预先设定孤立点个数,具有较高的准确率与检测效率,并且对数据集的分布状况不敏感。数据集上的实验结果表明,该算法能够高效、准确地识别孤立点。