Vol. 29 · No. 6

Clustering is a division of data into groups of similar objects. Aspect mining is a process that tries to identify crosscutting concerns in existing software systems. The goal is to refactor the existing systems to use aspect oriented programming, in order to make them easier to maintain and to evolve. The aim of this paper is to present a new hierarchical clustering based approach in aspect mining. For this purpose we propose HAC algorithm (Hierarchical Agglomerative Clustering in aspect mining). Clustering is used in order to identify crosscutting concerns. We evaluate the obtained results from the aspect mining point of view, based on two quality measures that we have previously introduced and a newly defined one. The proposed approach is compared with other similar existing approaches in aspect mining and two case studies are also reported.

Timed Chi ( chi) is a timed process algebra, designed for Modeling, simulation, verification and real-time control. Its application domain consists of large and complex manufacturing systems. The straightforward syntax and semantics are also highly suited to architects, engineers and researchers from the hardware design community. There are many different tools for timed Chi that support the analysis and manipulation of timed Chi specifications; and such tools are the results of software engineering research with a very strong foundation in formal theories/methods. Since timed Chi is a well-developed algebraic theory from the field of process algebras with timing, we have the idea that timed Chi is also well-suited for addressing various aspects of hardware systems (discrete-time systems by nature). To show that timed Chi is useful for the formal specification and analysis of hardware systems, we illustrate the use of timed Chi with several benchmark examples of hardware systems.

For a given regression model, each individual prediction may be more or less accurate. The average accuracy of the system cannot provide the error estimate for a single particular prediction, which could be used to correct the prediction to a more accurate value. We propose a method for correction of the regression predictions that is based on the sensitivity analysis approach. Using predictions, gained in sensitivity analysis procedure, we build a secondary regression predictor whose task is to predict the signed error of the prediction which was made using the original regression model. We test the proposed methodology using four regression models: locally weighted regression, linear regression, regression trees and neural networks. The results of our experiments indicate significant increase of prediction accuracy in more than 20

The resource allocation problem is the problem of assigning resources needed to accomplish some tasks. Such problems are present in many practical domains and are generally very hard to solve. In the paper two multi-agent systems for resource allocation are presented. Each of them uses different mechanisms (based on the ideas of ``free agents'' and ``life energy'') for maintaining functional integrity related with the number of agents in the population. The preliminary results of simulation experiments are also presented and the mechanisms of maintaining functional integrity are analyzed.

The paper addresses the problem of real valued detectors in ship immune system. The task of the system mentioned is to differentiate self objects, i.e. objects that are not dangerous to our ship, from other objects that can be a potential threat. To this end, mechanisms adapted from artificial immune systems are used. Since in the traditional model of artificial immune system binary strings are used to represent detectors and objects, in this paper modifications to this model are proposed. The modifications mentioned use real valued vectors instead of binary ones. To test the ship immune system equipped with real valued detectors, experiments were carried out. In the experiments, the task of the system was to differentiate self ship radio stations from non-self ones. Results of the experiments are presented at the end of the paper.

This paper presents a Field Programmable Gate Array (FPGA) implementation of a calculation module for exponential part of Gaussian Type Orbital (GTO). The module is composed of several specially crafted floating-point modules which are fully pipelined and optimized for high performance. The hardware implementation revealed significant speed-up for the finite sum of the exponential products calculation ranging from 2.5x to 20x in comparison to a general-purpose Central Processing Unit (CPU) version. Calculating values of GTOs is one of computationally critical parts of the Kohn-Sham algorithm. The approach proposed in the paper aims to increase the performance of a part of the quantum chemistry computational system by employing FPGA-based accelerator. Several issues are addressed, such as identification of code fragments which benefit most from hardware acceleration, porting a part of the Kohn-Sham algorithm to FPGA, data precision adjustment and data transfer overhead. The authors' intention was also to make hardware implementation of calculating the orbital function universal and easily attachable to different quantum-chemistry software packages.

Recently, a mixed-radix decimation in frequency (DIF) fast MDCT algorithm only for the mixed-radix decompositions or composite lengths N = 3m times 2, m > 0, has been proposed in citeShuBao. An improved mixed-radix DIF fast MDCT algorithm both in terms of the regularity and computational complexity is described. Based on observed simple algebraic identities in the original proposed algorithm citeShuBao, new formulas are derived resulting in a very regular computational structure. Consequently, the number of arithmetic operations is reduced significantly. Moreover, the improved algorithm is extended to all composite lengths N = 3m times 2p, m, p > 0. The improved algorithm defines new sparse matrix factorizations of the MDCT matrix for the composite lengths N = 3m times 2p, m, p > 0, and finally it provides new implementations of the forward/backward MDCT in MPEG-1/2 layer III (MP3) audio coding standard.

This paper presents a new approach to flood mapping using satellite synthetic-aperture radar (SAR) images that is based on intelligent tech- break niques. In particular, we apply artificial neural networks, self-organizing Kohonen's maps (SOMs), for SAR image segmentation and classification. Our approach was used to process data from different satellite SAR instruments (ERS-2/SAR, ENVISAT/ASAR, RADARSAT-1) for different flood events: the Tisza river, Ukraine and Hungary, 2001; the Huaihe river, China, 2007; the Mekong river, Thailand and Laos, 2008; and the Koshi river, India and Nepal, 2008.

Fault simulator for acpASC needs to deal with hazards, oscillations and races. The simplest algorithm for simulating faults is the serial fault simulation technique which was successfully used for the acpASC. Faster fault simulation techniques, for example deductive fault simulation, was previously used for the combinational and synchronous sequential circuits only. In this paper a deductive fault simulator for the stuck-at faults of acSI acpASC is presented. An algorithm for the propagation of the fault lists is proposed which can deal with the complex gates of the acpASC. The implemented deductive fault simulator was tested using acSI benchmark circuits. The experimental results show significant reduction of the computation time and negligible increase of the memory requirements in comparison with the serial fault simulation technique.