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.     

Interaction of graphene,MnOx, and Ca2+ for enhanced biomimetic, ‘bubble-free’ oxygen evolution reaction at mild pH

Water electrolysis powered by renewable electricity will likely be critical to a future hydrogen economy. However, the typical use of strongly acidic or alkaline electrolytes necessitates the use of expensive materials, while bubbles add to capital and operational costs, due to blocking of the electrode surface and the necessary use of pumps and gas-liquid separators. Here ‘bubble-free’ oxygen evolution at mild pH is carried out using an electrocatalyst that mimics photosystem II (PSII). The bubble-free electrode includes a gas-extracting Gore-Tex® membrane. Edge-functionalised graphene (EFG) is included to mimic the metal-binding local protein environment, and the tyrosine residue, in the oxygen evolving complex (OEC) of PSII, while MnO_x and Ca²⁺ are incorporated to mimic the Mn₄CaO₅ cluster. Interaction between EFG, MnO_x, and Ca²⁺ results in a significant, 130 mV fall in the overpotential required to drive electrocatalytic oxygen evolution at 10 mA cm⁻², compared to the electrode without these biomimetic components.     

多带抗噪声语音识别算法研究

根据Flether等人的研究,基于感知独立性假设的子带识别方法被用于抗噪声鲁棒语音识别。本文拓展子带方法,采用基于噪声污染假定的多带框架来减少噪声影响。论文不仅从理论上分析了噪声污染假定多带框架在识别性能上的潜在优势,而且提出了多带环境下的鲁棒语音识别算法。研究表明：多带框架不仅回避了独立感知假设要求,而且与子带方法相比,多带方法能更好的减少噪声影响,提高系统识别性能。     

Acquisition and pattern recognition of spectrum information of welding metal transfer

With the creation of a pattern recognition system for metal transfer mode, this article has collected five kinds of spectrum signals in gas metal arc welding (MIG, MAG and CO₂) and take them as training samples. These samples have been pretreated by computer, several key characteristic parameters of the spectrum signal have been creatively extracted, and a corresponding recognition function and a minimum-distance-classifier have been constructed. The results show that the pattern recognition of several kinds of metal transfer modes for the metal gas arc welding can be done successfully, and relative important parameters in welding process, such as the frequency of droplet transfer and the approximate diameter of each droplet, can also be obtained.     

基站智能天线多波束合成技术

采用数字波束形成的技术，给出了基站智能天线多波束合成的有效算法。对圆环阵分布的基站天线多波束方向图进行了综合，得到了空间隔离度很好的多个波束，并将副瓣控制在30dB以下，结果令人满意。该方法可用于任意单元构成的不均匀阵列天线，还可用于波束展宽及扇形波束的合成。     

基于Powell图的模式识别法建立化学反应速度方程

本文基于Ｐｏｗｅｌｌ图，提出了用模式识别技术中的相似系数法来确定化学反应级数的方法，并给出了计算实例。     

薄板坯连铸工艺液芯压下的有限元分析

简要介绍世界几种主要薄板坯连铸工艺。着重叙述应用大型通用商业软件MSC.MARC。模拟液芯压下时铸坯的变形，确定最佳压下方案的有限元分析，液芯压下应该在结晶器出口处及二冷区扇形段开始部分进行，这样能减小铸坯的应变。     

Pattern expressions and pattern automata

We define the pattern expressions as an extension of both regular expressions and patterns. We prove several properties of the new family of languages, similar to those of extended regex languages [Câmpeanu et al., Int. J. Found. Comput. Sci. 14 (6) (2003) 1007-1018]. We also define an automata system that recognizes these languages. Differences between regex and pattern expressions are also discussed.     

基于奇异值分解和神经网络的个性化信息模式识别

随着信息资源的持续增长，信息获取系统为特定用户提供特定信息服务的需求越来越重要。该文提出了基于奇异值分解和神经网络的个性化信息模式识别的方法，可以方便地获取用户信息需求的特征，并且利用这些特征信息完成个性化的信息识别和匹配，即实现个性化信息的模式识别，进而实现信息的个性化获取和管理。该文中的算法实现简单，可以方便地应用到目前的一些信息管理、服务系统中。