Feature Selection Based on Difference and Similitude in Data Mining

Feature selection is the pretreatment of data mining. Heuristic search algorithms are often used for this subject. Many heuristic search algorithms are based on discernibility matrices, which only consider the difference in information system. Because the similar characteristics are not revealed in discernibility matrix, the result may not be the simplest rules. Although differencesimilitude（DS） methods take both of the difference and the similitude into account, the existing search strategy will cause some important features to be ignored. An improved DS based algorithm is proposed to solve this problem in this paper. An attribute rank function, which considers both of the difference and similitude in feature selection, is defined in the improved algorithm. Experiments show that it is an effective algorithm, especially for large-scale databases. The time complexity of the algorithm is O（｜ C ｜^2｜U ｜^2）.     

Analysis of active sites for N2 and H^＋ reduction on FeMo-cofactor of nitrogenase

Dinitrogen (N2) and proton (H ),which act as physiological substrates of nitrogenase,are reduced on FeMo-co of the MoFe protein. However,researchers have different opinions about their exact reduction sites. Nitrogenases were purified from the wild type (WT) and five mutants of Azotobacter vinelandii (Av),including Qα191K,Hα195Q,nifV-,Qα191K/nifV- and Hα195Q/nifV-; and the activities of these en-zymes for N2 and H reduction were analyzed. Our results suggest that the Fe2 and Fe6,atoms closed to the central sulfur atom (S2B) within FeMo-co,are sites for N2 binding and reduction and the Mo atom of FeMo-co is the site for H reduction. Combining these data with further bioinformatical analysis,we propose that two parallel electron channels may exist between the 8Fe7S cluster and FeMo-co.     

数据挖掘的任务、对象和方法

数据挖掘是从大量的业务数据中抽取潜在知识的过程 ,是一个近年才在国内兴起的非常有实用价值的研究领域 .从数据挖掘的知识模式、任务、对象、方法、步骤给以综述 .     

一种基于Rough集的工程项目投标风险分析方法

针对工程项目投标过程中的不确定性和风险分析问题，从Rough集的基本体系出发，给出投标风险决策分析的知识表达系统，运用Rough集理论和方法对投标风险决策表的属性及属性值进行约简，挖掘其中隐含的知识，从而得出投标风险决策的最小决策算法，帮助决策者迅速对工程项目的投标风险进行评估和预测。     

Neuronal growth inhibitory factor inhibits pheochromo-cytoma PC12 in vitro

Neuronal growth inhibitory factor (GIF),named Metaliothioneins-Ⅲ (MT-Ⅲ), is the first protein validated to be capable of inhibiting the growth of neurons in nervous system. We have detected the effects of recombinant GIF on the growth of neuroblastoma SY5Y and pheochromocytoma PC12 by the MTT (Thiazolyl blue) reduction assay. Recombinant GIF inhibited PC12 in vitro; the inhibitory rate was about 25% when GIF was at 100 mg/L; and the inhibitory rate was about 50% when GIF was at 300 mg/L. It is shown that PC12 could serve as a proper model for detecting neuronal growth inhibitory activity of GIF. Recombinant GIF did not inhibit neuroblastoma SY5Y in vitro, a common model of neuroma; it is also shown that GIF could not inhibit neuromata extensively. The reason for GIF inhibiting PC12 may be that PC12 have some properties of cholinergic neuron. It must play an important role in discovering the mechanism of GIF's neuronal growth inhibitory activity.``     

文本分类实现技术

文本分类是文本数据挖掘的重要技术.从文本分类实现过程的各个环节,包括建立文档模型、特征提取、维数约简、选择分类策略几个方面分别给出了目前实用的解决方案,同时对各种算法进行了分类和性能上的定性与定量的比较,最后讨论了国内文本分类研究中的一些问题和未来的发展.     

高岭土碳热还原氮化制备Sialon粉末的反应过程研究

以高岭土为原料，采用碳热还原氮化法合成了Sialon粉末．通过对产物物相组成分析，探讨了合成Sialon粉末的反应过程，并对可能发生的不利于Sialon粉末合成的部分反应进行了热力学分析．研究表明：反应条件对产物组成有显著影响．氮气流量小，产物中有SiC生成；反应温度低、时间短将有莫来石相存在；温度过高或时间太长则产生A1N多型体和Si3N4相；只有足够的氮气流量、适宜的温度和反应时间才有利于Sialon的合成．     

制备方法对Ce-Ni/ Al2O3催化剂还原和吸氧性能的影响

采用程序升温还原和O2室温吸附方法,研究了两种方式制备的Ce Ni/Al2O3催化剂的还原行为和吸氧能力,以Ni/Al2O3和Ce/Al2O3为对照物研究了Ce、Ni的相互作用及其对催化剂的还原行为和吸氧能力的影响.实验发现,单独负载在Al2O3上的Ce和Ni组分的吸氧能力有很大差异,Ce物种表现出良好的吸氧能力;Ce和Ni组分同时负载在Al2O3上时,由于组分间的相互作用,促进了催化剂上低温可还原物种的形成,改变了低温可还原物种的分布,提高了623K还原后的吸氧能力;先浸Ce物种并分解后再浸Ni的Ce Ni/Al2O3催化剂,其吸氧能力增加更为显著.     

一种基于小波的白噪声消除方法

介绍了小波变换的基本理论，研究了信号和白噪声在小波变换下的不同特性，在此基础上提出了一种新的基于小波变换的白噪声消除方法．仿真结果证明，这种方法具有很好的去噪声效果．     

一种Web文档在线自适应分类新方法

Web文档自动分类是Web挖掘中的重要研究内容。文档向量空间模型(VSM)是实现文档自动分类的基础，但如何排除冗余属性并降低向量空间的维数是一个难点。文中运用粗集理论对由样本文档集合构成的信息系统进行数据泛化，并求取文档的最优规约属性集，大大降低了文档的特征空间的维数，减少了冗余属性对文档分类的干扰，提高了分类效率。运用Fuzzy ARTMAP(adaptive resonance theory mapping)神经网络，利用其自适应分类和增量学习的优良特性，实现了对Web文档的在线自适应分类。