Su-Shing Chen

Pool2: a generic system for cognitive map development and decision analysis

The authors present Pool2, a generic system for cognitive map development and decision analysis that is based on negative-positive-neutral (NPN) logics and NPN relations. NPN logics and relations are extensions of two-valued crisp logic, crisp (binary) relations, and fuzzy relations, NPN logics and relations assume logic values in the NPN interval (-1, 1) instead of values in (0, 1). A theorem is presented that provides conditions for the existence and uniqueness of heuristic transitive closures of an NPN relation. It is shown that NPN logic and NPN relations can be used directly to model a target world with a combination of NPN relationships of attributes and/or concepts for the purposes of cognitive map understanding, and decision analysis. Two algorithms are presented for heuristic transitive closure computation and for heuristic path searching, respectively. Basic ideas are illustrated by example. A comparison is made between this approach and others. >

A cognitive-map-based approach to the coordination of distributed cooperative agents

A partial taxonomy for cognitive maps is provided. The notions of NPN (negative-positive-neural) logic, NPN relations, coupled-type neurons, and coupled-type neural networks are introduced and used as a framework for cognitive map modeling. D-POOL a cognitive-map-based architecture for the coordination of distributed cooperative agents, is presented. D-POOL consists of a collection of distributed nodes. Each node is a cognitive-map-based metalevel system coupled with a local expert/database system (or agent). To solve a problem, a local node first pools cognitive maps from relevant agents in an NPN relation that retains both negative and positive assertions. New cognitive maps are then derived and focuses of attentions are generated. With the focuses, a solution is proposed by the local node and passed to the remote systems. The remote systems respond to the proposal, and D-POOL strives for a cooperative or compromised solution through coherent communication and perspective sharing. The utility of D-POOL is demonstrated using two examples in distributed group decision support. >

An intelligent distributed environment for active learning

Active learning is an effective learning approach. In this article we present an intelligent agent-assisted environment for active learning to better support the student-centered, selfpaced, and highly interactive learning approach. The environment uses the students learningrelated profile such as learning style and background knowledge in selecting, organizing, and presenting learning material, and it adopts a new approach to course content organization and delivery based on smart instructional components that can be integrated into a wide range of courses. The environment is being implemented using the prevalent Internet, Web, digital library, and multiagent technologies.

The paradox of digital preservation

Preserving digital information is plagued by short media life, obsolete hardware and software, slow read times of old media, and defunct Web sites. Herein lies the paradox: we want to maintain digital information intact, but we also want to access this information in a dynamic use context. Failure to address digital preservation problems is analogous to squandering potential professional, personal and economic gains, contributing to cultural and intellectual poverty and resulting in exorbitant costs for recovery. We are compelled to meet the research challenge to resolve the conflict between the creation context and the use context to facilitate digital information preservation.

Better learning for bidirectional associative memory

In this paper, we are to develop better learning rules for the bidirectional associative memory (BAM) based on three well-recognized optimal criteria, that is, all desired attractors should be made not only stable but also asymptotically stable, and spurious memories should be the fewest possible. We first explore the equivalence between the stability of all desired attractors and certain bidirectional linear separabilities and then relate three optimal criteria to expanding the kernal basin of attraction of each desired attractor in both the X-space and Y-space. To characterize this, we define an important and equivalent concept called by the Hamming-stability. Surprisingly, the Hamming-stability of all desired attractors turns out to be further equivalent to certain moderately expansive bidirectional linear separabilities. As a result, the well-known Rosenblatts perceptron learning rule can be used to achieve the stability of all desired attractors and even three optimal criteria. The learning rules developed thereofare called the bidirectional perceptron stability learning rule (BPSL) and bidirectional perceptron Hamming-stability learning rule (BPHSL). A number of computer experiments show the improved performance of the BAM trained by the BPSL or BPHSL in regard to its stability, asymptotic stability, and spurious memories.

Shape and motion of nonrigid bodies

Abstract In computer vision and image understanding research, rigidity of bodies and their motions has long been a key assumption in obtaining 3-dimensional shape and motion information from 2-dimensional images. In this paper, the rigidity assumption is relaxed in several fundamental problems. Nonrigid, elastic bodies under perspective projection are considered. Deformations of such bodies are diffeomorphisms of R3 in the case of static elasticity. For dynamic elasticity, generalized motions (including deformations) are represented by 1-parameter families of diffeomorphisms of R3. The primary goal is to determine a generalized motion of a known nonrigid, elastic body given images of the body before and after the generalized motion. Although there is no uniqueness result in general, a closed form solution is obtained for an isometric generalized motion given the corresponding image transformation. Further, an approach to the shape-from-shading method is presented which yields complete local surface information.

An agent enabling personalized learning in e-learning environments

In this work we focus on the potential of intelligent agents in e-learning environments. A central problem is the modeling of human learners so that the agent can facilitate personalized learning. The solution presented here is based on a connectionist approach. An intelligent agent, which replaces the human instructor, controls an e-learning environment and exploits a self-organizing map (SOM) learner behavioral model in order to achieve the learning goal. The learner SOM model has been implemented. The experiments support the argument that e-learning environments are feasible and can significantly assist dissemination of knowledge within modern educational settings.

Proteomics studies of traumatic brain injury.

Publisher Summary This chapter discusses the proteomics studies of traumatic brain injury. Traumatic brain injury (TBI) or traumatic head injury is characterized as a direct physical impact or trauma to the head causing brain injury. There is mechanical compression-induced direct tissue injury often associated with hemorrhage and contusion at the site of impact. The models of TBI include controlled cortical impact (CCI) model, a fluid percussion model and vertical weight drop models. The sample types that can be exploited for the proteomics analysis include brain tissues, cerebrospinal fluid (CSF), and blood (serum and plasma). The protein separation methods used for proteomic analysis are: (1) two-dimensional gel isoelectrofocusing/electrophoresis and (2) multidimensional liquid chromatography (LC). Proteomics approaches to identify and develop clinically useful biomarkers for brain injury from trauma, disease, or drugs involving protein separation by gels or LC, coupled with mass spectrometry, provide a potent and novel methodological array in detection of biomarkers of CNS injury either alone or in combination. The TBI proteomics core technologies will provide an integrative approach to genomic and proteomics information by developing a common portal architecture; the TBI proteomics portal.

A multi-agent system for computer science education

In this paper, we present a multi-agent system for supporting student-centered, self-paced, and highly interactive learning in undergraduate computer science education. The system is based on a hybrid problem-based and case-based learning model, for both creative problem solving and mechanical experience simulation. It aims at enhancing the effectiveness of the undergraduate learning experience in computer science. Implemented using the prevalent Internet, Web, and digital library technologies, the system adopts an open architecture design and targets at large-scale, distributed operations. In the initial implementation of the system, a number of prototypes using different Java-based software environments have been developed. They offer tradeoffs in system performance and design complexity.

Short communication: Specificity rule discovery in HIV-1 protease cleavage site analysis

Several machine learning algorithms have recently been applied to modeling the specificity of HIV-1 protease. The problem is challenging because of the three issues as follows: (1) datasets with high dimensionality and small number of samples could misguide classification modeling and its interpretation; (2) symbolic interpretation is desirable because it provides us insight to the specificity in the form of human-understandable rules, and thus helps us to design effective HIV inhibitors; (3) the interpretation should take into account complexity or dependency between positions in sequences. Therefore, it is necessary to investigate multivariate and feature-selective methods to model the specificity and to extract rules from the model. We have tested extensively various machine learning methods, and we have found that the combination of neural networks and decompositional approach can generate a set of effective rules. By validation to experimental results for the HIV-1 protease, the specificity rules outperform the ones generated by frequency-based, univariate or black-box methods.