Self-organization of behavioral primitives as multiple attractor dynamics: A robot experiment

This paper investigates how behavior primitives are self-organized in a neural network model utilizing a distributed representation scheme. The model is characterized by so-called parametric biases which adaptively modulate the encoding of different behavior patterns in a single recurrent neural net (RNN). Our experiments, using a real robot arm, showed that a set of end-point and oscillatory behavior patterns are learned by self-organizing fixed points and limit cycle dynamics that form behavior primitives. It was also found that diverse novel behavior patterns can be generated by modulating the parametric biases arbitrarily. Our analysis showed that such diversity in behavior generation emerges because a nonlinear map is self-organized between the space of parametric biases and that of the behavior patterns. The origin of the observed nonlinearity from the distributed representation is discussed. This paper investigates how behavior primitives are self-organized in a neural network model utilizing a distributed representation scheme. Our robot experiments showed that a set of end-point and oscillatory behavior patterns are learned by self-organizing fixed points and limit cycle dynamics that form behavior primitives. It was also found that diverse novel behavior patterns, in addition to previously learned patterns, can be generated by taking advantage of nonlinear effects that emerge from the distributed representation.     

Learning distributed representations of concepts using linearrelational embedding

We introduce linear relational embedding as a means of learning a distributed representation of concepts from data consisting of binary relations between these concepts. The key idea is to represent concepts as vectors, binary relations as matrices, and the operation of applying a relation to a concept as a matrix-vector multiplication that produces an approximation to the related concept. A representation for concepts and relations is learned by maximizing an appropriate discriminative goodness function using gradient ascent. On a task involving family relationships, learning is fast and leads to good generalization     

Using 3D printing technology with experiential learning strategies to improve preengineering students' comprehension of abstract scientific concepts and hands‐on ability

This study combined 3D printing technology with experiential learning strategies (ELS) to design a hands‐on curriculum for preengineering students. The participants learned interdisciplinary knowledge and abstract scientific concepts through the curriculum. The study implemented a quasi‐experimental design to examine whether the students who learned using the 3D printing technology with ELS demonstrated better learning performances regarding the comprehension of abstract scientific concepts and hands‐on ability. This study selected 184 10th‐grade students from five classes, which were divided into three groups. The experimental process was conducted over a period of 11 weeks (for a total duration of 960 min). It was found that all of the preengineering students improved their comprehension of abstract scientific concepts. The students who learned using the 3D printing technology understood abstract scientific concepts better than those who learned using the traditional hands‐on tools, and the students who learned using the 3D printing technology with ELS demonstrated better hands‐on ability than the other two groups. Using 3D printing technology with ELS resulted in significant positive effects on the participants' handmade processes, in which the students reinforced the connection between knowledge and handmade products, resulting in better comprehension of abstract scientific concepts and hands‐on ability.     

Multi-layer logic — A predicate logic including data structure as knowledge representation language

A new generation computer is expected to be the knowledge processing system of the future. However, many aspects are yet unknown regarding this technology, and a number of fundamental concepts, directly concerning knowledge processing system design need investigation, such as knowledge, data, inference, communication, information management, learning, and human interface. These concepts are closely related to knowledge representation. In particular, methodology to materialize such concepts as above in computers are completely dependent upon them. Thus, knowledge representation is a key concept in the design of knowledge processing systems and, consequently, of new generation computer systems. Knowledge representation design is a very important task affecting the performance of new generation computer systems to be developed. We should first investigate the requirements for precise knowledge representation, considering its effects on system performance, then design knowledge representations to satisfy these requirements. This paper discusses (1) a new style of information processing, (2) requirements for knowledge representation and (3) a knowledge representation satisfying these requirements, a knowledge processing system designed on this basis and a new style of problem solving using this system.     

Knowledge representation framework for interactive capture and management of reflection process in product concepts development

A design process can be characterized by reflection-in-action; that is, the process consists of a series of problem solving activities and each is embodied with a problem and a solution. In this process, a designer represents a hypothetical concept on each design alternative, deploys and verifies the concept from multiple viewpoints considering other alternatives, and modifies it. An advanced integrated design environment should be based on a representation framework that embodies this process of reflection in concept development, which usually remains in the realm of the designer’s tacit knowledge. This paper proposes a knowledge representation framework for an integrated design environment, named DRIFT (Design Representation Integration Framework of Three layers), which interactively captures and manages reflection processes of generating and verifying design concepts. The core of DRIFT is a three-layered design process model of actions, operations, and argumentation. This model integrates various design tools and captures performed design activities. The action level captures the sequence of design operations. The model operation level captures the transition of design states, recording a design snapshot over design tools, which are integrated through ontology-based representation of design concepts. The argumentation level captures the process of defining problems and corresponding alternative solutions. Integration of three levels with a template of design operation extracted from Design-For-X approaches enables a proposed system to interactively and efficiently capture and manage the process of design concept development through operations over design tools. A design operation template works to limit the number of links between the three levels remaining easy to manage its semantics. This paper also demonstrates a prototype implementation of DRIFT and its application to conceptual design of a small mechatronic system with a system modeling method. The paper concludes with a discussion of some future issues.     

基于本体的产品设计知识表示研究

针对目前的产品设计知识表示模型或是只局限于产品设计的某一阶段,或是局限于某一设计领域,对产品设计知识的概念存在着分类不清等不足,文中提出一种产品设计知识本体构建方法,并构建基于本体的产品设计知识表示模型。通过分析产品设计知识本体产生的知识概念、属性及知识概念间关系,形式化地定义和描述表示产品设计领域所包含的知识资源,提高企业产品设计知识共享和重用效率。以色选机设计知识表示为例,验证了所提出模型的有效性。     

Transferring Deep Object and Scene Representations for Event Recognition in Still Images

This paper addresses the problem of image-based event recognition by transferring deep representations learned from object and scene datasets. First we empirically investigate the correlation of the concepts of object, scene, and event, thus motivating our representation transfer methods. Based on this empirical study, we propose an iterative selection method to identify a subset of object and scene classes deemed most relevant for representation transfer. Afterwards, we develop three transfer techniques: (1) initialization-based transfer, (2) knowledge-based transfer, and (3) data-based transfer. These newly designed transfer techniques exploit multitask learning frameworks to incorporate extra knowledge from other networks or additional datasets into the fine-tuning procedure of event CNNs. These multitask learning frameworks turn out to be effective in reducing the effect of over-fitting and improving the generalization ability of the learned CNNs. We perform experiments on four event recognition benchmarks: the ChaLearn LAP Cultural Event Recognition dataset, the Web Image Dataset for Event Recognition, the UIUC Sports Event dataset, and the Photo Event Collection dataset. The experimental results show that our proposed algorithm successfully transfers object and scene representations towards the event dataset and achieves the current state-of-the-art performance on all considered datasets.     

Human experience and product usability: principles to assist the design of user-product interactions

This paper introduces research that investigates how human experience influences people's understandings of product usability. It describes an experiment that employs visual representation of concepts to elicit participants' ideas of a product's use. Results from the experiment lead to the identification of relationships between human experience, knowledge, and context-of-use - relationships that influence designers' and users' concepts of product usability. These relationships are translated into design principles that inform the design activity with respect to the aspects of experience that trigger people's understanding of a product's use. A design tool (ECEDT) is devised to aid designers in the application of these principles. This tool is then trialled in the context of a design task in order to verify applicability of the findings.     

SOIM: a self-organizing invertible map with applications in activevision

We propose a novel neural network, called the self-organized invertible map (SOIM), that is capable of learning many-to-one functionals mappings in a self-organized and online fashion. The design and performance of the SOIM are highlighted by learning a many-to-one functional mapping that exists in active vision for spatial representation of three-dimensional point targets. The learned spatial representation is invariant to changing camera configurations. The SOIM also possesses an invertible property that can be exploited for active vision. An efficient and experimentally feasible method was devised for learning this representation on a real active vision system. The proof of convergence during learning as well as conditions for invariance of the learned spatial representation are derived and then experimentally verified using the active vision system. We also demonstrate various active vision applications that benefit from the properties of the mapping learned by SOIM.     

基于构件的软件体系结构实现技术

软件体系结构设计给出了实现大规模软件复用,尤其是领域复用的重要设计模式。体系结构描述了软件系统的整体组织结构与风格,为基于构件的软件开发过程提供了构件组装的依据和上下文。文章提出了一种使用构件技术来构造软件构架的方法,使得软件构架不再是抽象意义上的概念模型,而是可以象构件一样使用的可以进行大规模软件复用的可执行代码。     

基于解释的算法构架的学习

本文提出了将解释学习方法用于学习算法构架的思想，以提高软件自动化系统从功能规格说明转换到设计规格说明的能力．文中给出了算法构架的表示，操作性的定义及其处理方法．系统从用户给出的一个问题的解中学习算法构架，用于解决一类问题，系统的学习效果表现为通过学习能够解决原来不能解的问题．     

THE WHAT,THE HOW,AND THE WHY IN DESIGN

Abstract

There seems to be confusion in the literature to date regarding the concepts of intent, purpose, function, behavior, and structure in design. This paper is meant to provide a focus for discussion on the disambiguation of such concepts. The paper argues for the integration of concepts relating to both the human sociocultural environment and the physical environment in arriving at definitions meaningful to designers from different disciplines. Issues such as decomposition and specialization are addressed, and finally, a means of representation with the expressive power and flexibility required is presented.     

On Developing Distributed Middleware Services for QoS- and Criticality-Based Resource Negotiation and Adaptation

The Global Resource Management System (GRMS) provides middleware services for QoS- and criticality-based resource negotiation and adaptation across heterogeneous computing nodes and communication networks. This paper presents GRMS's design, prototyping, and performance evaluation. We introduce GRMS design principles and two key concepts unified resource model and ripple scheduling--and describe our architectural design based on these concepts. Further, we present a decentralized end-to-end two-phase negotiation and adaptation protocol with the functionality of distributed, dynamic QoS adjustment and stream preemption. We discuss GRMS's system prototyping and lessons learned and report our experimentation and simulation results, providing insights into design and implementation of a middleware-based distributed resource management system.     

基于鲁棒子空间学习的粒子滤波跟踪算法*

线性子空间模型能够有效地描述目标表面受到光照和姿势变化的情况,然而大多数基于子空间表面模型的目标跟踪算法是在跟踪之前通过训练不同光照和姿势下目标的观测图像,得到一组特征基,并用这组特征基表示不同时刻目标表面变化,一旦训练完成之后,特征基就保持不变,不能在线更新。采用增量子空间学习的方法来构建目标表面的特征基,该特征基能够在线适应目标表面的变化。另一方面,传统的子空间学习方法是基于最小二乘重构误差,该方法容易受到异常测量数据的影响,为此采用鲁棒的子空间学习方法来降低异常测量数据对特征空间更新的影响。最后将鲁     

Software abstraction of elements of statistical strategy

This paper presents some general approaches to building software representations of statistical strategies. In statistics, strategy is the skilful management of data, probability models, experimental designs, and other statistical concepts. This paper addresses the representation of these concepts separately from the representation of the actions taken on them. The issue of having a credible visual interface with these representations is also raised.     

Mining fuzzy association rules from uncertain data

Association rule mining is an important data analysis method that can discover associations within data. There are numerous previous studies that focus on finding fuzzy association rules from precise and certain data. Unfortunately, real-world data tends to be uncertain due to human errors, instrument errors, recording errors, and so on. Therefore, a question arising immediately is how we can mine fuzzy association rules from uncertain data. To this end, this paper proposes a representation scheme to represent uncertain data. This representation is based on possibility distributions because the possibility theory establishes a close connection between the concepts of similarity and uncertainty, providing an excellent framework for handling uncertain data. Then, we develop an algorithm to mine fuzzy association rules from uncertain data represented by possibility distributions. Experimental results from the survey data show that the proposed approach can discover interesting and valuable patterns with high certainty.     

Learning Game-Specific Spatially-Oriented Heuristics

This paper describes an architecture that begins with enough general knowledge to play any board game as a novice, and then shifts its decision-making emphasis to learned, game-specific, spatially-oriented heuristics. From its playing experience, it acquires game-specific knowledge about both patterns and spatial concepts. The latter are proceduralized as learned, spatially-oriented heuristics. These heuristics represent a new level of feature aggregation that effectively focuses the program's attention. While training against an external expert, the program integrates these heuristics robustly. After training it exhibits both a new emphasis on spatially-oriented play and the ability to respond to novel situations in a spatially-oriented manner. This significantly improves performance against a variety of opponents. In addition, we address the issue of context on pattern learning. The procedures described here move toward learning spatially-oriented heuristics for autonomous programs in other spatial domains.