Abstract Associative classification is the integration of classification and association rule discovery in data mining. This approach often derives higher quality classifiers with reference to classification accuracy than traditional classification approaches such as decision trees and rule induction. In this paper, the problem of rule pruning in associative text categorisation is investigated. We propose five rule pruning methods within an existing... Associative classification is the integration of classification and association rule discovery in data mining. This approach often derives higher quality classifiers with reference to classification accuracy than traditional classification approaches such as decision trees and rule induction. In this paper, the problem of rule pruning in associative text categorisation is investigated. We propose five rule pruning methods within an existing associative classification algorithm called MCAR. Experimental results against large text collection (Reuters-21578) using the developed pruning methods as well as other known existing methods (Database coverage, lazy pruning) are conducted. The bases of the experiments are the classification accuracy and the number of generated rules. The results derived show that the proposed rule pruning methods derive higher quality and more scalable classifiers than those produced by lazy and database coverage pruning approaches. In addition, the numbers of rules generated by the developed pruning methods are usually less than those of lazy and database coverage. This makes the associative classifiers generated when using the proposed methods more maintainable by end users. Read More Read Less

Abstract The European construction industry faces a challenging environment with very tight time scales and budgets. At the same time, a lack of (real time) information makes it impossible to control projects effectively. The research project EPOS (Efficient process design by satellite supported software in the earth moving and road construction industry) which is described in this paper aims to alleviate... The European construction industry faces a challenging environment with very tight time scales and budgets. At the same time, a lack of (real time) information makes it impossible to control projects effectively. The research project EPOS (Efficient process design by satellite supported software in the earth moving and road construction industry) which is described in this paper aims to alleviate this. It is pointed out how this could be achieved by using intelligent performance and cost analysis of data from construction sites. At first the basic ideas of the EPOS project and the underlying corporate performance management approach are explained. Thereafter the system architecture and its data sources are briefly outlined. At a glance the main components, namely a machine guidance system, a production activity control (PAC) component, an enterprise resources planning (ERP) system, and a business intelligence (BI) software are introduced. An explanation of the performance and cost analysis components follows. Examples for data analysis are given. The final sections outline possibilities for further developments and summarize the paper. Read More Read Less

Abstract Mobile Ad hoc Network (MANET) is a collection of mobile nodes that form a dynamic topology and highly resource constrained network which are the most important characteristics of a MANET come from node mobility and Infrastructureless nature of MANETs. These characteristics essentially lead to adopt a routing protocol that quickly adapts to nodes mobility. Routing protocols in MANETs are classified... Mobile Ad hoc Network (MANET) is a collection of mobile nodes that form a dynamic topology and highly resource constrained network which are the most important characteristics of a MANET come from node mobility and Infrastructureless nature of MANETs. These characteristics essentially lead to adopt a routing protocol that quickly adapts to nodes mobility. Routing protocols in MANETs are classified into three types: proactive (table-driven), reactive (on-demand) and hybrid protocols. A routing protocol also may be considered a single path protocol or multipath protocol based on the routing mechanism. A routing protocol should try to minimize routing overhead such as control traffic and delay time overhead, maximize data packet delivery and throughput. This paper provides a comparative study for the performance of reactive against proactive routing protocols in MANETs. A comparison study has been introduced between proactive-reactive, single path-multipath protocols, and between reactive protocols only in terms of packet delivery fraction, average end-to-end delay, routing overhead and throughput. This evaluation and comparison study have been performed by means of simulation using NS2. Read More Read Less

Abstract An approach to rectify the distortions caused by the Travelling Wave Tube (TWT) or solid-state High Power Amplifiers (HPA) used in digital satellite communication systems and at the same time to correct the effects of intersymbol interference is presented. Linear adaptive signal processing techniques fail to eliminate these nonlinear distortions introduced by nonlinear components in the satellite channel. Here we... An approach to rectify the distortions caused by the Travelling Wave Tube (TWT) or solid-state High Power Amplifiers (HPA) used in digital satellite communication systems and at the same time to correct the effects of intersymbol interference is presented. Linear adaptive signal processing techniques fail to eliminate these nonlinear distortions introduced by nonlinear components in the satellite channel. Here we present a nonlinear adaptive signal processing technique based on Volterra series. First part of this paper [15] includes mathematical description of proposed scheme. This part includes computer simulation and results obtained. Read More Read Less

Abstract The work reported in this paper aims to formalise two approaches based on intelligent agents and intelligent cores which are employed for automated fault tolerance in high-performance computing systems. A layered cognitive architecture is therefore mapped onto intelligent agents and intelligent cores, specifically taking into account the perceiving, reasoning, judging, responding and learning capabilities of agents and cores for demonstrating... The work reported in this paper aims to formalise two approaches based on intelligent agents and intelligent cores which are employed for automated fault tolerance in high-performance computing systems. A layered cognitive architecture is therefore mapped onto intelligent agents and intelligent cores, specifically taking into account the perceiving, reasoning, judging, responding and learning capabilities of agents and cores for demonstrating their cognitive intelligence. The mapping results in three views of cognition, namely an agent’s view, a swarm’s view and a landscape’s view of cognition, referred to as ‘triad views’. A qualitative evaluation of the mapping on the basis of a set of reference criteria is presented. Read More Read Less