Shih-Hsi Liu

Replication and comparison of computational experiments in applied evolutionary computing: Common pitfalls and guidelines to avoid them

Abstract Replicating and comparing computational experiments in applied evolutionary computing may sound like a trivial task. Unfortunately, it is not so. Namely, many papers do not document experimental settings in sufficient detail, and hence replication of experiments is almost impossible. Additionally, some work fails to satisfy the thumb rules for Experimentation throughout all disciplines, such that all experiments should be conducted and compared under the same or stricter conditions. Also, because of the stochastic properties inherent in evolutionary algorithms (EAs), experimental results should always be rich enough with respect to Statistics. Moreover, the comparisons conducted should be based on suitable performance measures and show the statistical significance of one approach over others. Otherwise, the derived conclusions may fail to have scientific merits. The primary objective of this paper is to offer some preliminary guidelines and reminders for assisting researchers to conduct any replications and comparisons of computational experiments when solving practical problems, by the use of EAs in the future. The common pitfalls are explained, that solve economic load dispatch problems using EAs from concrete examples found in some papers.

QoS-UniFrame: a Petri net-based modeling approach to assure QoS requirements of distributed real-time and embedded systems

Assuring quality of service (QoS) requirements is critical when assembling a distributed real-time and embedded (DRE) system from a repository of existing components. This paper presents a two-level approach for assuring satisfaction of QoS requirements in the context of a reduced design space for DRE systems. A dynamic and parallel approach is introduced to prune off the infeasible design spaces at the first level. Evolutionary algorithms cooperating with a domain-specific scripting language then discard less probable design spaces using statistics. These techniques fulfill the collective objectives of pruning and assuring the design space at system assembly time.

A Semantics- and Data-Driven SOA for Biomedical Multimedia Systems

Due to the problems of heterogeneous data and platforms, abundant functional and QoS requirements, high data size, and tangling correlation between data/contents and software functionalities, developing large-scale biomedical multimedia database systems is a challenging task. This paper presents a semantics- and data-driven service-oriented architecture (SOA) to take the interoperability and scalability advantages of conventional SOA and solve the aforementioned problems. By establishing data ontology with respect to data properties, contents, QoS, and biomedical regulations and expanding service ontology to describe more functional and QoS specifications supported by services, appropriate services for processing biomedical multimedia data may be discovered, performed, tuned up or replaced as needed. Additionally, six transmission services are introduced to support dynamic adaptation under specific requirements.

A non-invasive approach to dynamic Web services provisioning

Service-oriented computing has emerged as a new component-based software development paradigm in a network-centric environment. By using a standard description language and protocol, services can be used to wrap legacy software systems to be integrated beyond the enterprise boundary across heterogeneous platforms. Nevertheless, the challenges come in tandem with the opportunities because of the inherent dynamic characteristics within a distributed environment. In particular, there is a need for dynamic adaptation for provisioned services to accommodate the ever-changing business requirements externally as well as the computing resource status internally, while maintaining the continuousness of service provisioning. We present a dynamic Web service provisioning approach based on .NET common language runtime, one of the two primary Web services platforms, exploring the runtime code manipulation at the intermediate language (IL) level rather than at the source code level. Meanwhile, we show how the service provisioning can be adapted in a modularized way by complementing the conventional service-oriented architecture (SOA) with a repository of adaptation aspects. Moreover, we demonstrate how dynamic service provisioning can be used for nonfunctional property assurance.

On the Importance of the Artificial Bee Colony Control Parameter ‘limit’

Artificial Bee Colony (ABC) is a successful meta-heuristic algorithm that has been greatly utilised by researchers. Through our practical experience of ABC, we have noticed that the recommended formula ‘limit’ = ne * D may not be the best choice for different problems. In this work, a set of experiments using horizontal and vertical approaches has been designed and executed with the aim of observing the effect of ‘limit’ on ABC. The results have been statistical analysed using Null Hypothesis Significance Testing (NHST) as well as the Chess Rating System for Evolutionary Algorithms (CRS4EAs), which is a novel approach for comparing meta-heuristic algorithms. It is shown that the recommended formula is not the best setting for different problems and approaches. Hence, the control parameter ‘limit’ should be tuned or controlled. The other important result of this study is to show that CRS4EAs is comparable but also shows benefits over NHST.DOI: http://dx.doi.org/10.5755/j01.itc.46.4.18215

A clustering entropy-driven approach for exploring and exploiting noisy functions

Linear, Gaussian, fitness proportional, clustering, and Rosca entropies are succinct measures of diversity that have been applied to balance exploration and exploitation in evolutionary algorithms. In previous studies, an entropy-driven approach using linear entropy explicitly balances and/or searches optimal solutions for the selected unimodal and multimodal functions excluding noisy functions. This paper investigates the reasons for such an exception and introduces a clustering entropy-driven approach to solve the problem. Such an approach provides a coarse-grained diversity measure that filters the noise of functions, varies cluster size and categorizes individuals at the genotype level. The experimental results show that the clustering entropy-driven approach further improves the searching results of noisy functions by one more degree.

A Semantics and Data-Driven Biomedical Multimedia Software System

Developing a large-scale biomedical multimedia software system is always a challenging task: Satisfaction of sundry and stringent biomedical multimedia related requirements and standards; Heterogeneous software deployment and communication environments; and tangling correlation between data/contents and software functionalities, among others. This paper introduces a novel biomedical multimedia software system developed under Service-Oriented Architecture (SOA). Such a system takes the advantage of interoperability of SOA to solve the heterogeneity and correlation problems. The paper also classifies the system into services, annotation, ontologies, semantics matching, and QoS optimization aspects which may potentially solve the requirements problem: By establishing data ontology with respect to data properties, contents, QoS, and biomedical regulations and expanding service ontology to describe more functional and QoS specifications supported by services, appropriate services for processing biomedical multimedia data may be discovered, performed, tuned up or replaced as needed. Lastly, a biomedical education project that improves the performance of feature extraction and classification processed afterwards is introduced to illustrate the advantages of our software system developed under SOA.

Enhanced Web Service Matchmaking: A Quality of Service Approach

With the increasing number of Web services, the Semantic Web research community is moving toward enhancing procedures so that clients may effectively discover appropriate Web services that maximally satisfy not only functional but also non-functional requirements. Using purely syntactic approaches to discover Web services has limited efficiency. Therefore, semantic matchmaking registries can play an important role in providing better results. However, most of the semantic registries make decisions based on only functional requirements. Due to the proliferation of Web services that have similar functionality, there is a need to further filter Web services according to Quality of Service (QoS) specifications. This work introduces a rich QoS ontology inspired from the World Wide Web Consortium that defines relationships among QoS attributes. The QoS attributes defined here are mainly used to define network related characteristics. However, the work can be extended by defining QoS ontologies for other domains. Additionally, we extend the test collection OWLS-TC according to the QoS ontology. The extension is intended to integrate quantified values for QoS attributes into Web service descriptions. A matchmaking algorithm based on similarity measurements is presented. Moreover, deterministic and adaptive parameter control techniques are introduced to guide the requester, when needed, to adjust QoS specifications so better similarity can be achieved. Parameter control algorithms integrate Pellet, an OWL reasoner, to reason about dependencies among QoS concepts in the QoS ontology.

T-Clipse: an integrated development environment for two-level grammar

T-Clipse is an Integrated Development Environment (IDE) using Eclipse developed for Two-Level Grammar (TLG), a high-level formal specification language. In our research, given a programming language, TLG is used to specify the syntax and semantics of the language to generate an interpreter for that language. This IDE provides a user-friendly environment for editing and navigation of TLG source codes, as well as parsing and error tracing at the source code level and interpreter generation for programming languages. To build this IDE, we leveraged the extension power of plug-ins in Eclipse as well as a TLG parser prebuilt from an existing lexical analyzer generator and parser generator.

Introducing domain-specific language implementation using web service-oriented technologies

Several advantages have been documented that suggest Domain-Specific Languages DSLs have the potential to improve productivity, reliability, maintainability and portability in some specialized domains. However, several key challenges still remain. In particular, the extension and evolution of both DSL syntax and semantics still suffer due to the limitations related to the current state-of-the-art implementation techniques. Such techniques also lack interoperable capabilities among base languages and limited tool support. As changes of domain concepts are omnipresent and more base languages may support DSL implementation, the aforementioned limitations may be no longer tolerable, and hence a new implementation technique to DSL development is desired. This paper implements six DSL case studies representing imperative, declarative and hybrid categories to validate the feasibility of utilizing Service-Oriented Architecture SOA for DSL implementation. Such case studies also highlight that the advantages of SOA i.e., ease of evolution/extension, interoperability and tool support can be retained under the context of DSL development. The paper concludes with the discussion of additional findings, both positive and negative: the SOA-based approach improves modularization at the lexical, syntactical and semantic levels and delegates tokenization/parsing to the underlying WS-BPEL engine; yet, the usability, resource utilization, security, and flexibility of the SOA-based DSLs are degraded, which requires more future work in this unique area that spans SOA and DSLs.