融合零样本学习和小样本学习的弱监督学习方法综述

深度学习模型严重依赖于大量人工标注的数据,使得其在数据缺乏的特殊领域内应用严重受限。面对数据缺乏等现实挑战,很多学者针对数据依赖小的弱监督学习方法开展研究,出现了小样本学习、零样本学习等典型研究方向。对此,本文主要介绍了弱监督学习方法条件下的小样本学习和零样本学习,包括问题定义、当前主流方法以及实验设计方案,并对典型模型的分类性能进行对比。然后,给出零-小样本学习的问题描述,总结研究现状和实验设计,并对比典型方法的性能。最后,基于当前研究中出现的问题对未来研究方向进行展望,包括多种弱监督学习方法的融合与理论基础的探究,以及在其他领域的应用。     

融合零样本学习和小样本学习的弱监督学习方法综述

深度学习模型严重依赖于大量人工标注的数据，使得其在数据缺乏的特殊领域内应用严重受限。面对数据缺乏等现实挑战，很多学者针对数据依赖小的弱监督学习方法开展研究，出现了小样本学习、零样本学习等典型研究方向。对此，本文主要介绍了弱监督学习方法条件下的小样本学习和零样本学习，包括问题定义、当前主流方法以及实验设计方案，并对典型模型的分类性能进行对比。然后，给出零-小样本学习的问题描述，总结研究现状和实验设计，并对比典型方法的性能。最后，基于当前研究中出现的问题对未来研究方向进行展望，包括多种弱监督学习方法的融合与理论基础的探究，以及在其他领域的应用。     

基于深度学习和集成学习的辐射源信号识别

随着电磁环境的日益复杂和雷达辐射源信号类型的逐渐增多,如何有效地识别雷达信号类型成为一个重要的问题。为解决这个问题,提出了一种基于深度学习和集成学习的辐射源信号识别框架。该框架由特征提取和分类器设计两部分组成。第一部分,将雷达信号变换到时频域,利用栈式降噪自编码模型学习时频图像的特征。深度模型的训练采用无监督预学习和有监督微调相结合。第二部分,构造一个集成不同支持向量机分类器的模型对雷达信号进行识别。利用8种不同的辐射源信号验证了提出模型的有效性,结果表明结合这两种机器学习的方法有助于提高辐射源信号的识别正确率。     

基于案例学习和目标驱动学习的观点生成框架

群体观点生成是观点评价和群体一致性意见达成的基础．本文提出基于案例学习和目标驱动学习的观点生成框架,将人工智能技术集成入群体决策支持系统中,通过动态问题分解过程、多案例经验学习过程、主动的目标驱动学习过程的协调运作,既支持初始观点生成,又支持观点的修改与扩充．本文采用描述性智能辅助技术,使得观点生成的智能辅助方法不仅易实现,而且具有高度的灵活性和适应性．     

带有初态学习的指数变增益迭代学习控制

针对一类非线性时变系统在有限时间区间上的轨迹跟踪问题,提出一种新的迭代学习控制算法,该算法对系统的控制输入和初始状态同时采用闭环指数变增益迭代学习律。基于算子理论,对具有任意初始状态的系统,在该迭代学习律作用下的收敛性进行严格证明,同时给出该迭代学习算法收敛的谱半径形式的充分条件。该算法与固定增益的迭代学习控制相比较,不仅加快了收敛速度,而且还解决了指数变增益迭代学习控制要求初始状态严格重复的问题。仿真结果表明了该算法的有效性。     

自学习智能决策支持系统

提出了一种基于神经网络与专家系统的自学习智能决策支持系统,使得智能决策支持系统能够对过去的经验进行学习。在无人作战飞机系统的应用表明,智能决策支持系统不仅能够对过去的经验进行学习,并且成功地解决了在规则不完善、战场信息不完全的情况下决策支持系统的推理问题。仿真结果表明,提出的方法显著地提高了智能决策支持系统的性能。     

神经网络自适应学习研究

本文讨论了网络拓扑结构、网络的学习参数以及神经元的激活函数等多方面的人工神经网络学习问题，提出了具体实现方法。实验表明这些方法对于加快网络的收敛速度，优化网络的拓扑结构等方面有显著成效。     

在线学习中基于群体智能的学习路径推荐方法

研究在线学习中学习路径的推荐问题.在线学习中的推荐系统可以自动识别学习者的个性化需求,主动推荐符合用户需求的学习路径,从而提高在线学习系统的个性化服务水平.对此,提出了一种扩展蚁群算法用于解决学习路径的推荐问题,该算法在进行推荐决策时综合参考了学习者群体对学习路径的评价以及目标用户在知识水平和学习风格上的特点.模拟实验...     

递归神经网络学习速率研究

针对典型的对角递归神经网络,推导出递归神经网络稳定条件下网络输出层、隐含层及关联层学习速率的具体取值范围。提出设计者可通过从具体系统中获得的数据确定网络各层学习速率的上、下界数值,确定自适应学习速率的初值与调节方法,并可选取最优学习速率。因而该法具有很强的可操作性和实用性。还给出一个具体数值实例,说明自适应学习速率与最佳学习速率的调整过程。     

Biased Learning Creates Overconfidence

The aim of this paper is to develop a multi-period economic model to interpret how the people become overconfident by a biased learning that people tend to attribute the success to their abilities and failures to other factors. The authors suppose that the informed trader does not know the distribution of the precision of his private signal and updates his belief on the distribution of the precision of his knowledge by Bayer’s rule. The informed trader can eventually recognize the value of the precision of his knowledge after an enough long time biased learning, but the value is overestimated which leads him to be overconfident. Furthermore, based on the definition on the luckier trader who succeeds the same times but has the larger variance of the knowledge, the authors find that the luckier the informed trader is, the more overconfident he will be; the smaller the biased learning factor is, the more overconfident the informed trader is. The authors also obtain a linear equilibrium which can explain some anomalies in financial markets, such as the high observed trading volume and excess volatility.     

Leadership and Learning: A Critical Reexamination of Senge’s Learning Organization

From its inception the concept of the learning organization has been identified with a particular type of organization or new forms of organizational learning. But it is often forgotten that Senge’s ‘system thinking’ formulation of the learning organization was inseparable from an attempt to reformulate a new way of thinking about change agency and leadership in organizations. Here it is argued that Senge’s learning organization can be re-conceptualised as a partial fusion of ‘systems thinking’ and learning theories that leads to a concept of organizational learning as a form of ‘distributed leadership’. However, the concept is critically flawed because it cannot theorise the organizing practices by which learning to lead and leading to learn are shared or distributed in organizations. It is concluded that Senge’s under-theorized focus on distributed leadership consistently neglects issues of practice and issues of power. As such his work does not provide an exploration of the possibilities for increasing the dispersal of human agency, power, knowledge and autonomy within the workplace.     

From Teaching to Learning: Technological Potential and Sustainable, Supported Open Learning

The central theme of this paper is the current interest in most educational institutions in moving from teaching to learning as their main system model and the implications which technology media have for unravelling the debate and influencing the resulting practice (for example see Active Learning: Using the Internet for Teaching, Number 2, July, 1995), Our chosen strategy for dealing with the central theme is to consider how we use language, metaphor and models to describe systems for teaching and learning and what is the role of technology in these systems. In particular we describe how the Open University is moving from an analysis of individual technologies to a synthesis of the educational ideas into a sustainable system that conforms to the University's policy of supported open learning. This includes focusing on open and equal access to courses, considerable attention to staff development and training (specifically in core teaching areas such as face to face tuition, correspondence tuition, student support, telephone and other media supported communication) as well as the necessity and value of reflection on practice (e.g. see Baker, Tomlinson et al,, 1996). In this first section, taking as our starting point the traditional linear view of educational structures, we build upon the notion of the learning system and describe this as an approach with a somewhat long and surprising history originating in the Socratic method. An eductive, cyclic learning model is introduced, and the historical impact of technology on this model is briefly reviewed in a global context by addressing the specific issue of access from the developing countries. Following from this, in sections two and three two types of teaching are described, under the labels of conventional and distance. It is argued that each has strengths and tendencies towards the eductive learning system introduced in section one. However, it is also argued that each has flaws which provide problems for the development of a learning system which can be sustained at distance. An analysis based upon metaphor is applied. In section four, it is then argued that in the conventional model there is a system with excellent learning potential but with an organisational structure which often does not encourage it whereas in the distance model there is the potential organisation for providing feedback but often little effective use made of it. That lack of use includes the new technologies which are now being focused on by the Open University, both in terms of the INSTILL initiative (Integrate New Systems and Technologies Into Lifelong Learning) and the Technology Strategy for Academic Advantage (Laurillard, Christmas et al., 1996). These technologies are argued, in section five, to be capable of being tactically effective in stimulating and supporting the strategic aim of learning. Technology mediation is discussed both in terms of course-based teaching and in terms of research opportunities. In section six a range of problem/opportunity areas of immediate concern for the extension of this technologically mediated system to the developing countries are discussed, while policy implications for the extension of supported open learning are drawn out in section seven. These include policies toward co-learning, access to learning, quality standards and the authorship of educational material.     

The Learning Control and Learning Adaptive Control of General Nonlinear Systems：MIMO Case

ＴｈｅＬｅａｒｎｉｎｇＣｏｎｔｒｏｌａｎｄＬｅａｒｎｉｎｇＡｄａｐｔｉｖｅＣｏｎｔｒｏｌｏｆＧｅｎｅｒａｌＮｏｎｌｉｎｅａｒＳｙｓｔｅｍｓ：ＭＩＭＯＣａｓｅＨＯＵＺｈｏｎｇｓｈｅｎｇａｎｄＨＡＮＺｈｉｇａｎｇ（ＩｎｓｔｉｔｕｔｅｏｆＡｐｐｌｉｅｄＭａ...     

Enhancing Collaborative Learning in a Blended Learning Environment: Applying a Process Planning Model

The research reported here was undertaken in a blended learning environment where contact tuition is “blended” with computer-supported (online) learning. The context is a developing country (South Africa). After completion of two full cycles of an action inquiry the planning for the third cycle was undertaken according to an existing process planning model which was originally developed for the design and planning of team-based action learning and action research (ALAR) projects. The illustration of the application of the model focuses on the aspect of student collaboration in a blended learning environment and shows how the original process planning model was adapted and applied during the re-planning phase of the project. The final deliverable is a set of action plans for future collaborative learning that could strengthen student-centred learning experiences in a context which still bears the scars of an unfortunate segregated past.     

On Learning Adaptive Service Compositions

Service composition is an important and effective technique that enables atomic services to be combined together to form a more powerful service,i.e.,a composit...