Browsing by Author "Jiao, L."

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Multiobjective optimization of classifiers by means of 3-D convex hull based evolutionary algorithms
    (Elsevier, 2016-05-27) Zhao, J.; Basto-Fernandes, V.; Jiao, L.; Yevseyeva, Iryna; Maulana, A.; Li, R.; Back, T.; Emmerich, M. T. M.; Tang, Ke
    The receiver operating characteristic (ROC) and detection error tradeoff(DET) curves are frequently used in the machine learning community to analyze the performance of binary classifiers. Recently, the convex-hull-based multiobjective genetic programming algorithm was proposed and successfully applied to maximize the convex hull area for binary classifi- cation problems by minimizing false positive rate and maximizing true positive rate at the same time using indicator-based evolutionary algorithms. The area under the ROC curve was used for the performance assessment and to guide the search. Here we extend this re- search and propose two major advancements: Firstly we formulate the algorithm in detec- tion error tradeoffspace, minimizing false positives and false negatives, with the advantage that misclassification cost tradeoffcan be assessed directly. Secondly, we add complexity as an objective function, which gives rise to a 3D objective space (as opposed to a 2D pre- vious ROC space). A domain specific performance indicator for 3D Pareto front approxima- tions, the volume above DET surface, is introduced, and used to guide the indicator-based evolutionary algorithm to find optimal approximation sets. We assess the performance of the new algorithm on designed theoretical problems with different geometries of Pareto fronts and DET surfaces, and two application-oriented benchmarks: (1) Designing spam filters with low numbers of false rejects, false accepts, and low computational cost us- ing rule ensembles, and (2) finding sparse neural networks for binary classification of test data from the UCI machine learning benchmark. The results show a high performance of the new algorithm as compared to conventional methods for multicriteria optimization.
  • No Thumbnail Available
    ItemMetadata only
    Supporting Provenance in Service-oriented Computing Using the Semantic Web Technologies
    (IEEE, 2006-12) Chen, Liming; Jiao, L.
    The Web is evolving from a global information space to a collaborative problem solving environment in which services (resources) are dynamically discovered and composed into workflows for problem solving, and later disbanded. This gives rise to an increasing demand for provenance, which enables users to trace how a particular result has been arrived at by identifying the resources, configurations and execution settings. In this paper we analyse the nature of service-oriented computing and define a new conception called augmented provenance. Augmented provenance enhances conventional provenance data with extensive metadata and semantics, thus enabling large scale resource sharing and deep reuse. A Semantic Web Service (SWS) based, hybrid approach is proposed for the creation and management of augmented provenance in which semantic annotation is used to generate semantic provenance data and the database management system is used for execution data management. We present a general architecture for the approach and discuss mechanisms for modeling, capturing, recording and querying augmented provenance data. The approach has been applied to a real world application in which tools and GUIs are developed to facilitate provenance management and exploitation.