Computing and Informatics (E-Journal - Institute of Informatics, SAS, Bratislava)
Not a member yet
1506 research outputs found
Sort by
Time Series Trend Analysis Based on K-Means and Support Vector Machine
Get PDFIn this paper, we apply both supervised and unsupervised machine learning techniques to predict the trend of financial time series based on trading rules. These techniques are K-means for clustering the similar group of data and support vector machine for training and testing historical data to perform a one-day-ahead trend prediction. To evaluate the method, we compare the proposed method with traditional back-propagation neural network and a standalone support vector machine. In addition, to implement this combination method, we use the financial time series data obtained from Yahoo Finance website and the experimental results also validate the effectiveness of the method
Extensible Host Language for Domain-Specific Languages
Get PDFProgramming languages greatly influence the way how programs are created and evolved. This means that the use of appropriate language for solved problem can greatly increase developer productivity. Composition of languages can provide great help in construction of a new language from existing components and for integration of several languages that may be needed to effectively solve a complex problem. In this paper we analyze the composition problem on the two levels: composition of languages and composition of concepts in a language. Possibilities of transition from language composition to concepts composition are also presented. Based on that, we propose a framework of languages construction based on concept composition that aims to support reusability of language elements and tools. It uses common host syntax for developed languages. Their semantics is defined in a general-purpose language. Proposed approach is demonstrated on example languages developed using prototype implementation
Handover Architectures for Heterogeneous Networks Using the Media Independent Information Handover (MIH)
Get PDFIn heterogeneous networks, network selection by nature is a multi-dimensional problem. Many parameters need to be considered for handover decision making. Apart from handover accuracy and efficiency, an important consideration is the scalability and signaling overhead of such handover algorithms. In this article we propose to break down a Simple Additive Weighting (SAW) based heterogeneous handover algorithm in two parts. The execution of the first part is carried out in an independent and proactive manner prior to the actual handover, assuming three different handover architectures. The handover architectures are differentiated based upon the level of the distribution of the handover algorithm among multiple network components. The Media Independent Handover (MIH) and its different services are used to retrieve and share information among MIH enabled nodes and for conformity among heterogeneous network standards. The proposed algorithm is evaluated with respect to handover accuracy, handover delay efficiency and signaling overhead. The evaluation is carried out for all three handover architectures using simulations. Only handovers between Wi-Fi (IEEE 802.11) and WiMAX (IEEE 802.16) networks are considered. But the handover framework is general and can be extended to consider other wireless and mobile communication networks
Combining the Continuous Integration Practice and the Model-Driven Engineering Approach
Get PDFThe software development approach called model-driven engineering has become increasingly widespread. The continuous integration practice has also been gaining the importance. Some works have shown that both can improve the software development process. The problem is that the model-driven engineering is still a very active research topic lacking its maturity, what translates into difficulties in optimal incorporation of the continuous integration practice in the process. We present an experience report in which we show the problems we have detected in a real project and how we have solved them. Thus, we increase the productivity of development and the non-technical people are able to modify already deployed applications. Finally, we incorporate an evaluation that shows the benefits of the proposed union
Analysis of Successful Trades Information Disclosure Mechanism
Get PDFIn most of e-commerce sites evaluation systems are employed to evaluate each user after trades. Normally, seller's evaluation score is shown in e-commerce site and computed based on the score given by buyers. In recent e-commerce sites, the evaluation is based on multiple attributes and buyers give their thinking for each attribute. In this viewpoint, the seller evaluation in e-commerce is collected knowledge from buyers and the synthetic score is computational intelligence because each buyer makes his/her decision whether he/she trades with the seller. In this paper, we discuss the computational intelligence of the evaluation system in e-commerce. Then, to avoid asymmetric and incomplete information in trades, we design a mechanism of the trader evaluation. After that, we present some experiments to show the rate of successful trades in our proposed mechanism. Contributions of this paper are showing the theoretical discussion of e-commerce computational intelligence, design of evaluation mechanism and the effectiveness of the proposed mechanism
Improvements on Gabor Descriptor Retrieval for Patch Detection
Get PDFThe localization of object parts in the component-based object detection is among the main tasks to solve. This paper presents several improvements of the proposed local image descriptor based on Gabor wavelets. Including these descriptors in the desired application is an ambitious challenge if we take into account the high number of parameters. Determining of parameters can be very hard because of their infinite definition range. Defining the filters is done in two stages: a theoretical consideration narrows the domain and the cardinality of parameters; this is followed by adequate experiments to select the most characteristic descriptor for a target image patch. The descriptor is created from a given number of 2D Gabor filters chosen by the GentleBoost learning algorithm. Comparing the proposed descriptor to those found in the state of the art, we can conclude that the selected filters are adaptable to any target object. In contrast to this, the majority of filter-based descriptors have fixed values for the parameters that do not allow to be ductile to the given object. Parameters fine-tuning allows the descriptor to be general, and discriminative at the same time. The effect of the following experiments has been analyzed during the investigation: elimination of redundancy between the weak classifiers, using the LoG interest points in the detection process. Finally, we propose an acceleration algorithm in order to deter- mine the response map faster. By means of the descriptor, the response map is created, which accurately localizes the target object part and can easily be integrated in almost all detection systems
PAFSV: A Formal Framework for Specification and Analysis of SystemVerilog
Get PDFWe develop a process algebraic framework PAFSV for the formal specification and analysis of IEEE 1800TM SystemVerilog designs. The formal semantics of PAFSV is defined by means of deduction rules that associate a time transition system with a PAFSV process. A set of properties of PAFSV is presented for a notion of bisimilarity. PAFSV may be regarded as the formal language of a significant subset of IEEE 1800TM SystemVerilog. To show that PAFSV is useful for the formal specification and analysis of IEEE 1800TM SystemVerilog designs, we illustrate the use of PAFSV with a multiplexer, a synchronous reset D flip-flop and an arbiter
Aspect-Oriented Formal Modeling: (AspectZ + Object-Z) = OOAspectZ
Get PDFThe aspect-oriented software development (AOSD) paradigm permits modularizing crosscutting concerns of base modules, a non-usual task in other software development paradigms. Since AOSD was born in the programming stage as an extension of an object-oriented (OO) programming language, and AOSD considers, in addition to base modules, new modules named aspects, then a complete AOSD process requires that each stage considers the base and aspect modules. Therefore, looking for an AOSD process, mainly to apply AOSD in other phases of the OO software development process, so far, different OO modeling tools and language extensions to support AOSD have been proposed. As an example, AspectZ is an extension of the formal language Z to support AOSD. To reach a transparency of concepts and design in AOSD, the main contribution of this article is to propose OOAspectZ, a formal language for the requirements specification stage of aspect-oriented (AO) software applications, that, firstly, extends AspectZ and, secondly, integrates Object-Z and AspectZ formal specifications. Thus, OOAspectZ supports relevant AO elements such as join points, and This and Target objects for join point events. As an application example, this article applies OOAspectZ to a system named GradUTalca for a Chilean university. For GradUTalca, this article presents AO UML use cases and UML class diagrams, formal Object-Z and OOAspectZ specifications, and a final woven specification to show an integration of Object-Z and OOAspectZ specifications
Evolutionarily Tuned Generalized Pseudo-Inverse in Linear Discriminant Analysis
Get PDFLinear Discriminant Analysis (LDA) and the related Fisher's linear discriminant are very important techniques used for classification and for dimensionality reduction. A certain complication occurs in applying these methods to real data. We have to estimate the class means and common covariance matrix, which are not known. A problem arises if the number of features exceeds the number of observations. In this case the estimate of the covariance matrix does not have full rank, and so cannot be inverted. There are a number of ways to deal with this problem. In our previous paper, we proposed improving LDA in this area, and we presented a new approach which uses a generalization of the Moore-Penrose (MP) pseudo-inverse to remove this weakness. However, for data sets with a larger number of features, our method was computationally too slow to achieve good results. Now we propose a model selection method with a genetic algorithm to solve this problem. Experimental results on different data sets demonstrate that the improvement is efficient
GPGPU Computing for Microscopic Simulations of Crowd Dynamics
Get PDFWe compare GPGPU implementations of two popular models of crowd dynamics. Specifically, we consider a continuous social force model, based on differential equations (molecular dynamics) and a discrete social distances model based on non-homogeneous cellular automata. For comparative purposes both models have been implemented in two versions: on the one hand using GPGPU technology, on the other hand using CPU only. We compare some significant characteristics of each model, for example: performance, memory consumption and issues of visualization. We also propose and test some possibilities for tuning the proposed algorithms for efficient GPU computations