多分类器的一种动态联合方法

多分类器联合是解决复杂模式识别问题的有效办法。对于多分类器联合,一个关键的问题是如何对每个分类器的分类性能作出可靠性估计。以往提出的方法是利用各个分类器在训练阶段得到的知识来判断决策的可靠性,这些方法都需要大量的存储空间,并且没有考虑到分类器在分类过程中,由于输入样本的质量变化从而分类性能也会改变。文章提出了一种分类器的动态联合方法,该方法直接利用分类器的输出信息来估计分类器的可靠性。实验结果表明,比较传统的联合方法,该方法是一种有效的联合方法。     

一种模糊分类器集成的方法

提出了一种基于模糊积分的模糊分类器集成的方法，该方法能在模糊分类器生成过程中，进一步减少主观因素的参与成份，使分类模器具有更好的稳定性和更高的分类识别率。给出了基于隶属度矩阵的模糊积分密度确定方法，介绍了基于模糊积分的分类器集成算法。用权威的数据集作为实验数据集，将提出方法与已有的分类器集成方法进行实验比较，评测了所提出方法的有效性。     

多分类器选择集成方法

针对目前人们对分类性能的高要求和多分类器集成实现的复杂性,从基分类器准确率和基分类器间差异性两方面出发,提出了一种新的多分类器选择集成算法。该算法首先从生成的基分类器中选择出分类准确率较高的,然后利用分类器差异性度量来选择差异性大的高性能基分类器,在分类器集成之前先对分类器集进行选择获得新的分类器集。在UCI数据库上的实验结果证明,该方法优于bagging方法,取得了很好的分类识别效果。     

一种简单有效的多分类器综合方法

针对小字符集脱机手写体汉字识别中的多分类器集成问题，提出了一种简单有效的综合方法，实验表明综合后系统的识别率明显高于单个分类器的识别率。     

基于置信度的多分类器互补集成手写数字识别

离线手写数字识别是光学字符识别的一个重要分支,在银行票据识别、邮政编码识别等领域有着广泛的应用。由于单一分类器在识别率上很难达到要求,人们提出了各种集成分类器识别方案。通过对离线手写数字的特征提取,从特征互补的角度出发,采用了最小距离分类器、树分类器和BP网络分类器进行多分类器互补集成,提出了基于置信度的多分类器互补集成方法。通过实验对比,基于置信度的多分类器互补集成手写数字识别在识别率和识别速度上达到了满意的结果。     

一种分类器选择方法

在按照“测试-选择”方法设计多分类器系统时,从超量生成的候选分类器集中选取一个最优子集是关键环节之一。基于此,定义一个组合适宜度概念,提出一种新的分类器选择方法。将该方法用于高光谱遥感数据分类实验中,并从具有27个候选的分类器集中挑选子集。实验结果表明,该方法在选择效率和识别精度方面具有优势,能保证所选子集的泛化能力。     

一种识别手写汉字的多分类器集成方法

根据多信源信息处理与字符识别的经验知识，提出了一个识别手写汉字的多分类器线性集成模型．这个模型不仅考虑到不同的分类器对不同字符识别能力的不同，而且还考虑了不同的分类器得出的输入字符与参考模板之间相似度的实际大小对判决的影响，及不同分类器提供的候选字符对判决的支持作用，更重要的是提供了一种通过监督学习，利用计算机程序自动计算模型参数的方法，因而实现了一个较好的集成系统．同时，本文还提供了三个用于集成的分类器，它们集成的结果充分显示了本方法的有效性。     

一种识别手写汉字的多分类器集成方法

根据多信源信息处理与字符识别的经验知识,提出了一个识别手写汉字的多分类器线性集成模型.这个模型不仅考虑到不同的分类器对不同字符识别能力的不同,而且还考虑了不同的分类器得出的输入字符与参考模板之间相似度的实际大小对判决的影响,及不同分类器提供的候选字符对判决的支持作用,更重要的是提供了一种通过监督学习,利用计算机程序自动计算模型参数的方法,因而实现了一个较好的集成系统.同时,本文还提供了三个用于集成的分类器,它们集成的结果充分显示了本方法的有效性.     

一种多分类器选择性集成的网络流量分类方法

针对多分类器集成方法产生的流量分类器在泛化能力方面的局限性,提出一种选择性集成网络流量分类框架,以满足流量分类对分类器高效的需求。基于此框架,提出一种多分类器选择性集成的网络流量分类方法 MCSE(Multiple Classifiers Selective Ensemble network traffic classification method),解决多分类器的选取问题。该方法首先利用半监督学习技术提升基分类器的精度,然后改进不一致性度量方法对分类器差异性的度量策略,降低多分类器集成方法实现网络流量分类的复杂性,有效减少选择最优分类器的计算开销。实验表明,与Bagging算法和GASEN算法相比,MCSE方法能更充分利用基分类器间的互补性,具有更高效的流量分类性能。     

基于置信度的手写体数字识别多分类器动态组合

多分类器组合利用不同分类器、不同特征之间的互补性，提高了组合分类器的识别率。传统的组合方法里，各分类器在组合中所承担的角色是固定的，而实际应用中，对于不同的测试样本，每个分类器识别结果的可信度是不同的。该文根据分类器置信度理论，提出了各类别的置信度。用测试样本自身的置信度信息实现分类器的动态组合，并把这种动态组合方法具体应用到手写体数字的识别。这种方法还可以在不影响已有数据的情况下添加新的分类器进行组合。     

Constructivist Management of Knowledge,Communication and Enterprise Innovation: Lessons from Indian Experience

Prime movers of enterprise innovation are inside the organisation. The enterprise NIIT, studied in this paper, tells us how elicitation and recognition of knowledge and its contents set an enterprise on the move. Recognised knowledge is actionable. Each act of recognition is an addition to knowledge content, and such additions took place in NIIT through acts of communications and self-searching. These contents form a large collage that cannot be strewn into a single novelistic episode. Perspectives and context, motives and suggestions render to each member of the enterprise different disjoint appearances of knowledge. There are thus many knowledge systems and multiple narrations, each with small episode-like finality. Enterprise innovation is achieved not through any grand episodic integration of all knowledge contents, but by constructively eliciting further new contents of knowledge, and encouraging acts of communications and discourses on this new knowledge. Constructive management of knowledge and communication towards enterprise innovation has thus been defined. Through such constructive management, NIIT enabled itself innovationally and empowered its members in eliciting knowledge and acting communicatively. A critical feature of knowledge work is that it requires multidisciplinary expertise and mutual learning in order to achieve a complex synthesis of highly specialised state-of-the-art technologies and knowledge domains. A convivial work culture and a culture of communicative acts enable sharing of the non-informatised yet recognised contents of knowledge.     

Regional Specialisation for Technological Innovation in R&D Laboratories: A Strategic Perspective

The present paper attempts to highlight the strategy of regional specialisation for technological innovation in R&D laboratories. The paper makes a proposition that regional specialisation should be recognised as a strategic initiative for technology development in R&D laboratories. The rationale for this strategic initiative has been substantiated with the help of illustrations from the cases of technology development efforts taken up in different laboratories in the country under the Council of Scientific and Industrial Research (CSIR), India. In this direction, CSIR and other centres of excellence have played a pioneering role in the development of various industrial clusters and artisan concentrations in different parts of the country. The implications of adoption or otherwise of this strategy initiative for technological innovation in R&D laboratories have been discussed.     

Driving Force for Sustainable Development: Principles of Harmony and Balance

This research focuses on the driving forces of technology development and the interactive relationships between technology and the factors that promote economic development and social progress. It aims at presenting a basic theory for sustainable development, as well as a foundation for decision-makers for drawing up an integrative strategy. As such it is an attempt at how to create harmonious relations between human progress, technology, economy, and society.     

Learning Organisations: The Process of Innovation and Technological Change

In the present scenario of globalisation, knowledge has become the prime factor of production for competitive advantage. This calls for acquisition and utilisation of knowledge for innovation and technical change on a constant basis, which is only possible in a ‘learning organisation’. Innovative activities of a learning organisation are influenced by three main factors: (1) internal learning; (2) external learning; and (3) the innovation strategies decided upon by the enterprise management. An assumption has been made that, particularly in developing countries, absorption and adaptation of technologies, i.e. indigenisation, take place through a process of ‘learning by doing’. Taking this into consideration, this paper focuses on a few case studies carried out at NISTADS, New Delhi, India, on small enterprises in the formal as well as traditional sectors, highlighting the learning process in an organisational context and how it brings in innovation and technological change at enterprise level. The study demonstrates that the learning environment in an organisational context is an indispensable process to be innovative and building up capabilities for technological change. This in turn also calls for strong networking of the enterprises with academia, R&D institutions and other enterprises, to create knowledge clusters. This builds up a strong case for a network approach of learning organisations not only at the regional level but also at the cross-cultural level for constant innovation and technical change.