Classification of data with imbalanced class distribution has posed a significant drawback of the performance attainable by most standard classifier learning algorithms, which assume a relatively balanced class distri...
详细信息
Classification of data with imbalanced class distribution has posed a significant drawback of the performance attainable by most standard classifier learning algorithms, which assume a relatively balanced class distribution and equal misclassification costs. This learning difficulty attracts a lot of research interests. Most efforts concentrate on bi-class problems. However, bi-class is not the only scenario where the class imbalance problem prevails. Reported solutions for bi-class applications are not applicable to multi-class problems. In this paper, we develop a cost-sensitive boosting algorithm to improve the classification performance of imbalanced data involving multiple classes. One barrier of applying the cost-sensitive boosting algorithm to the imbalanced data is that the cost matrix is often unavailable for a problem domain. To solve this problem, we apply Genetic Algorithm to search the optimum cost setup of each class. Empirical tests show that the proposed cost-sensitive boosting algorithm improves the classification performances of imbalanced data sets significantly.
Context awareness is an inherent feature of pervasive computing. It enhances the proactiveness of the system thus requiring less user attention and fewer human-machine interactions, it supports intelligent personaliza...
详细信息
Context awareness is an inherent feature of pervasive computing. It enhances the proactiveness of the system thus requiring less user attention and fewer human-machine interactions, it supports intelligent personalization features, and it can assist the system to address the user requirements considering the current conditions. Nevertheless, in such environments, various types of context information are involved and need to be efficiently managed and maintained, soundly interpreted, rapidly processed, and securely disseminated by the system. Thus, an interoperable and flexible context representation scheme is necessary that will support efficient context interpretation and reasoning and will perform well in distributed large-scale context-aware systems. This paper is concerned with the development of a hybrid context representation scheme that aims to combine the maintenance, distribution and administrative facilities of a location-based context model and the semantic advantages of context ontologies
This study explores the relationship between program slices and faults. The aim is to investigate whether the characteristics of program slices can be used to identify faultprone software components. Slicing metrics a...
The change in the softwaresystems is inevitable if the system is to remain useful. The change in business requirements sometimes is such that the old and new requirements have to be met simultaneously, i.e. old and n...
详细信息
The change in the softwaresystems is inevitable if the system is to remain useful. The change in business requirements sometimes is such that the old and new requirements have to be met simultaneously, i.e. old and new business rules are needed to exist concurrently. Accommodating more than one rule in the existing entities, leads to the change in the operational system, which involves risk. The evolution in the system impacts the business rules as well as the data model of the system. The evolution step incorporating the change in business rule requires changes to be carried out in the application architecture in terms of addition or modification of process as well as data components and also the alterations in the relationships among the components. The evolution problem of coexistence of business rules can be considered as incremental change to the existing architectural design of the system to nullify the risk of change to the software. In this paper we propose the framework, where temporal aspects of process as well as data components and concurrent validity of multiple business rules, with multiple versions of the components can be effectively implemented. The framework supports evolution without disturbing existing architecture and functionality. This framework is consistent in application to systems irrespective of the design methodology followed for the same.
Connectors are used to realize component interactions in component systems. Apart from their primary function, which is mediating the communication, their implementation can also support additional features that, whil...
详细信息
This paper assesses and analyzes a harmonic vector modeling method for reducing disturbance torque and motion state estimation error using quantized input measurements of position. There is a finite resolution availab...
详细信息
This paper assesses and analyzes a harmonic vector modeling method for reducing disturbance torque and motion state estimation error using quantized input measurements of position. There is a finite resolution available for any digital interface and quantization will always be present in some form. Although not possible to completely reconstruct a quantized measured state, it is possible to attenuate deterministic quantization characteristics. The discussed method models quantized position feedback as a spatially rotating vector consisting of fundamental and harmonic components. The error mitigation strategy involves decoupling these spatial quantization harmonics from the actual fundamental position, using heterodyning tracking observer topologies
Some applicants, developers, and commercial-off-the-shelf (COTS) software vendors have proposed reverse engineering as an approach for satisfying RTCA/DO-178B objectives for airborne software. RTCA/DO-178B, software C...
详细信息
Some applicants, developers, and commercial-off-the-shelf (COTS) software vendors have proposed reverse engineering as an approach for satisfying RTCA/DO-178B objectives for airborne software. RTCA/DO-178B, software Considerations in Airborne systems and Equipment Certification, serves as the means of compliance for most airborne software in civil aircraft. DO-178B defines reverse engineering as "the method of extracting software design information from the source code" and provides guidance particular to reverse engineering, when it is used to upgrade a development baseline. For purposes of this paper, reverse engineering is an approach for creating software life cycle data that did not originally exist, cannot be found, is not adequate, or is not available to a developer in order to meet applicable DO-178B objectives. Reverse engineering is not just the generation of data-rather it is a process to assure that the data is correct, the software functionality is understood and well documented, and the software functions as intended and required by the system. Reverse engineering is not, as some software developers propose, just an effort to generate the software life cycle data without intent to build in quality and the resulting design assurance. This paper explores reverse engineering in airborne software projects, by explaining a definition for the certification domain, describing the motivation for its use, and documenting the certification concerns. Two actual cases of reverse engineering are also described to illustrate the certification concerns in real projects.
Context awareness is an essential cornerstone in future pervasive computing systems. It has the potential to greatly reduce the user attention and interaction bottlenecks, to give humans the impression that services f...
详细信息
Context awareness is an essential cornerstone in future pervasive computing systems. It has the potential to greatly reduce the user attention and interaction bottlenecks, to give humans the impression that services fade into the background, and to support intelligent personalization features. Nevertheless, in order to create such an environment, a growing amount of personal information has to be provided to the system, either manually or automatically. Hence the digital trace and representation users have in the system is getting dangerously detailed, thus stressing the need for privacy protection. DAIDALOS (Satyanarayanan, 2001) is a European research project in the area of 3G and beyond, which aims to combine heterogeneous networks in a transparent and seamless way, and develop on top of this a pervasive environment for applications and end-users. This paper describes the main models and mechanisms that have been established to provide federated context-aware services and protect the privacy of their users
暂无评论