An intelligent web-page classifier with fair feature-subset selection

The explosion of on-line information has given rise to many manually constructed topic hierarchies (such as Yahoo!!). But with the current growth rate in the amount of information, manual classification in topic hierarchies results in an immense information bottleneck. Therefore, developing an automatic classifier is an urgent need. However, classifiers suffer from enormous dimensionality, since the dimensionality is determined by the number of distinct keywords in a document corpus. More seriously, most classifiers are either working slowly or they are constructed subjectively without any learning ability. In this paper, we address these problems with a fair feature-subset selection (FFSS) algorithm and an adaptive fuzzy learning network (AFLN) for classification. The FFSS algorithm is used to reduce the enormous dimensionality. It not only gives fair treatment to each category but also has ability to identify useful features, including both positive and negative features. On the other hand, the AFLN provides extremely fast learning ability to model the uncertain behavior for classification so as to correct the fuzzy matrix automatically. Experimental results show that both FFSS algorithm and the AFLN lead to a significant improvement in document classification, compared to alternative approaches.     

基于聚类算法实现信号盲分类

本文解决了信号处理、工业控制等领域存在的非平稳信号盲分类问题。在聚类中广泛应用的K-Means算法及其它基于中心的聚类算法有两个共同的缺陷-需要预先确定类数目且随机初始化中心引起性能不稳定。本文提出的算法较好地解决了这两个问题，提高了算法稳定性，实现了非平稳信号盲分类。提取非平稳信号的小波系数作为聚类的样本空间，分析聚类结果的统计偏差以估计类的数目，采用调和均值准则进行分类。最后给出的仿真结果表明本文提出的方法较传统的K-Means算法明显降低分类错误率。     

裂缝性油气藏采收率：100个裂缝性油气田实例的经验总结

通过对世界上100个裂缝性油气藏的综合评价，研究储集层及流体本身的性质(包括孔隙度、渗透率、黏度、可动油比例、含水饱和度、润湿性及裂缝分布特征等)和驱动机制及油藏管理战略(优化日产量和采用不同类型的提高采收率技术)对其最终采收率的影响。将裂缝性油气藏分为4类：I类的基质几乎没有孔隙度和渗透率，裂缝是储存空间和流体流动的通道；Ⅱ类的基质有较低的孔隙度和渗透率，基质提供储存空间，裂缝提供流动通道；Ⅲ类(微孔隙)的基质具有高孔隙度和低渗透率，基质提供储存空间，裂缝提供流动通道；Ⅳ类(大孔隙)的基质具有高的孔隙度和渗透率，基质提供储存空间和流动通道，裂缝仅增加渗透率。对26个Ⅱ类油气藏和20个Ⅲ类油气藏的开采历史的研究表明：Ⅱ类油气藏的采收率受水驱强度和最优日产量控制，日产量过高会很容易破坏Ⅱ类油气藏，一些Ⅱ类油气藏如果管理得当，采收率可以很高，不需要二次或三次采油；Ⅲ类油气藏的采收率主要受岩石和流体本身性质的影响，特别是基质渗透率、流体重度、润湿性以及裂缝强度等，不进行二次或三次采油不可能完全开采，往往需要采用一些提高采收率的专门技术。以往将Ⅱ类和Ⅲ类裂缝性油气藏归为一类，认识它们的区别将有助于选择更好的开发策略。     

Using classification techniques for informal requirements in the requirements analysis-supporting system

In order to efficiently develop large-scale and complicated software, it is important for system engineers to correctly understand users’ requirements. Most requirements in large-scale projects are collected from various stakeholders located in various regions, and they are generally written in natural language. Therefore, the initial collected requirements must be classified into various topics prior to analysis phases in order to be usable as input in several requirements analysis methods. If this classification process is manually done by analysts, it becomes a time-consuming task. To solve this problem, we propose a new bootstrapping method which can automatically classify requirements sentences into each topic category using only topic words as the representative of the analysts’ views. The proposed method is verified through experiments using two requirements data sets: one written in English and the other in Korean. The significant performances were achieved in the experiments: the 84.28 and 87.91 F1 scores for the English and Korean data sets, respectively. As a result, the proposed method can provide an effective function for an Internet-based requirements analysis-supporting system so as to efficiently gather and analyze requirements from various and distributed stakeholders by using the Internet.     

CIMSCLASS─—面向CIM的零件分类软件开发工具

介绍了一个能为工厂设置零件编码系统和开发零件分类系统的工具系统。该系统能适应用户的具体环境，能对零件的特征信息做较详细的描述和建立相应目标的零件分类系统，从而能在ＧＴ（成组技术）基础上为机械加工零件开发ＣＡＤ／ＣＡＰＰ／ＣＡＭ集成系统提供支持，以满足ＣＩＭＳ的需要。     

基于神经网络的VQ图象编码

文章给出将神经网络与矢量量化相结合的两种编码方法，介绍了其原理和实现的算法。经计算机模拟表明，在比特率低于０．３ｂｐｐ时仍可得到高质量的编码图象。     

对绿色技术内涵与类别的再思考

在分析不同国际组织、机构对环境友好技术概念的不同界定方法及其特点的基础上,以系统性、层次性和可操作性为原则,以环境亲和程度为标准,从基础绿色技术、创新绿色技术和生态绿色技术三个层次对绿色技术的涵义和类别进行了界定。     

Satellite SAR Remote Sensing of Great Lakes Ice Cover,Part 2. Ice Classification and Mapping

During the 1997 winter season, shipborne polarimetric backscatter measurements of Great Lakes (freshwater) ice types using the Jet Propulsion Laboratory C-band scatterometer, together with surface-based ice physical characterization measurements and environmental parameters, were acquired concurrently with Earth Resource Satellite 2 (ERS-2) and RADARSAT Synthetic Aperture Radar (SAR) data. This polarimetric data set, composed of over 20 variations of different ice types measured at incident angles from 0° to 60° for all polarizations, was processed to radar cross-section to establish a library of signatures (look-up table) for different ice types. The library is used in the computer classification of calibrated satellite SAR data. Computer analysis of ERS-2 and RADARSAT ScanSAR images of Great Lakes ice cover using a supervised classification technique indicates that different ice types in the ice cover can be identified and mapped, and that wind speed and direction can have an influence on the classification of water as ice based on single frequency, single polarization data. Once satellite SAR data are classified into ice types, the ice map provides important and necessary input for environmental protection and management, ice control and ice breaking operations, and ice forecasting and modeling efforts.     

Prediction of fire classification for wood based products. A multivariate statistical approach based on the cone calorimeter

Wood has long traditions as a building material, and is often used in construction elements, and as interior and exterior surfaces in the Nordic countries. In most applications, there are reaction to fire requirements to products used as surfaces, e.g. in escape routes and larger public spaces. Most wood products will therefore have to be treated with fire retardant (FR) agents to fulfil the strict requirements to properties connected to heat release and flame spread. Unfortunately, FR agents usually also increase the smoke production, as they cause a more incomplete combustion of the wood. The wood product manufacturers seek to find the optimal amount of FR additives where both heat release and smoke production in the classifying test are within the requirements given in the building regulations. This paper describes models for prediction of the European reaction to fire classes of wood products. The models are based on multivariate statistical analysis, and use test results from the cone calorimeter test as input. The presented models are, with very good precision, able to predict which Euroclass and additional smoke class a wood based product would obtain if it were to be tested in the single burning item test. Copyright © 2006 John Wiley & Sons, Ltd.     

Two-tier architecture for unconstrained handwritten character recognition

In this paper, we propose an approach that combines the unsupervised and supervised learning techniques for unconstrained handwritten numeral recognition. This approach uses the Kohonen self-organizing neural network for data classification in the first stage and the learning vector quantization (LVQ) model in the second stage to improve classification accuracy. The combined architecture performs better than the Kohonen self-organizing map alone. In the proposed approach, the collection of centroids at different phases of training plays a vital role in the performance of the recognition system. Four experiments have been conducted and experimental results show that the collection of centroids in the middle of the training gives high performance in terms of speed and accuracy. The systems developed also resolve the confusion between handwritten numerals.