[show abstract][hide abstract] ABSTRACT: A merger occurs when a bidder firm offers to purchase the control rights in a target firm or when a target firm solicits bids from a bidder firm to purchase the control rights. Predicting merger candidacy is important to measure the price impact of mergers. This study investigates the performance of artificial neural networks and multinomial logit models in predicting merger candidacy. We use a comprehensive dataset that covers the years 1979 to 2004 and includes all deals with publicly listed bidders and targets. We find that both models perform similarly while predicting target and non-merger firms. The multinomial logit model performs slightly better in predicting bidder firms.
[show abstract][hide abstract] ABSTRACT: This paper addresses the application of the principles of feedback and self-controlling software to the tabu search algorithm. We introduce two new reaction strategies for the tabu search algorithm. The first strategy treats the tabu search algorithm as a target system to be controlled and uses a control-theoretic approach to adjust the algorithm parameters that affect search intensification. The second strategy is a flexible diversification strategy which can adjust the algorithm’s parameters based on the search history. These two strategies, combined with tabu search, form the Self Controlling Tabu Search (SC-Tabu) algorithm. The algorithm is implemented and tested on the Quadratic Assignment Problem (QAP). The results show that the self-controlling features of the algorithm make it possible to achieve good performance on different types of QAP instances.
[show abstract][hide abstract] ABSTRACT: In this paper the principle of self adaptation is applied to achieve a self controlling software. The software considered in this case is a heuristic search algorithm: the reactive tabu search. In reactive search algorithms, the behavior of the algorithm is evaluated and modified during the search. To improve self adaptation, two new strategies for reactive tabu search are introduced. The first strategy uses a control theoretic approach, treats the algorithm as a plant to be controlled and modifies the algorithm parameters to control the intensification of the search. The second strategy adjusts several parameters according to the feedback coming from the search to achieve diversification during the search. These strategies adjust the parameters of the tabu search and form the self controlling tabu search (SC-Tabu) algorithm. The performance of the algorithm is tested on different problem types of the quadratic assignment problem (QAP). The results show that the algorithm adapts successfully to achieve good performance on problems with different structures.
Second IEEE International Conference on Self-Adaptive and Self-Organizing Systems, SASO 2008, 20-24 October 2008, Venice, Italy; 01/2008
[show abstract][hide abstract] ABSTRACT: In this paper we consider the impact of the speed of the processor on the stability of the adaptation process. The problem is investigated on a case study that involves a real-time system for monitoring illuminations of multiple emitters by one receiver mounted on a moving platform. This is a self-controlling system whose goal is to schedule the receiver so that the uncertainty about the times of illuminations of radars in the direction of the platform is minimized. The system has several controllers and a scheduler which are using the same computer as the computation resource. This paper demonstrates the effects of computation time on a real time self-controlling system and shows the changes in stability of the system under high computation loads.
12th IEEE International Conference on the Engineering of Computer-Based Systems (ECBS 2005), 4-7 April 2005, Greenbelt, MD, USA; 01/2005
[show abstract][hide abstract] ABSTRACT: In this research we examine the supply and demand relationship in a carbon emissions trading market by using agent based system simulation. An agent based system is a computational system that simulates the behaviors of autonomous agents. We model different nations which take place in the carbon trading market as autonomous agents, and these agents act as buyers or sellers in the same trading market. A market-director agent directs the trading between the buyer and seller agents and enables a market mechanism with dynamic auction. This study analyzes the changes in carbon price and the total carbon reduction amount under several different demand scenarios. We used JADE (Java Based Agent Development Environment) as the simulation environment. This paper presents an application of agent based simulation to analyze the carbon trading market. Agent based simulation is a tool to model and analyze the behavior of complex systems composed of autonomous agents. Agents act in an environment, which has policies and interaction protocols that design the communication between agents. Agents may have different behaviors, and strategic goals to achieve. Agents may learn throughout their lifetime, use their knowledge to forecast future events and adapt their behavior accordingly. Energy management related systems involve different interacting components, which can be modeled using multi agent systems and the behavior of these systems under different scenarios can be analyzed using agent based simulation. Kyoto Protocol aims to reduce the greenhouse gas emissions to prevent the climate change. Kyoto Protocol Annex-1 countries are expected to develop clean manufacturing processes to make this reduction. Kyoto protocol also brings three flexibility mechanisms to help countries meet their reduction commitments. ET (Emission Trading) is one of these mechanisms, and it helps Annex 1 regions meet their emission reduction target at a reduced cost (13). One type of emissions trading is carbon emission trading which is specifically for carbon dioxide. A nation buying carbon in the carbon trading market buys the right to burn it and a nation selling it gives up the right to burn it (14). Regions such as Former Soviet Union and Eastern Europe have surplus carbon credits called "hot air" because their emissions are already under the amounts of emissions specified in accordance with the Kyoto Protocol. This paper aims to analyze the supply and demand relationship in the carbon trading market, using agent based simulation approach. The paper is organized as follows: Section 2 presents a short review of recent agent based applications and gives background information on carbon trading. Section 3 presents the carbon trading model and
[show abstract][hide abstract] ABSTRACT: In this study, a semi-Markovian insurance model and a semi-Markovian random walk process (X(t)) which describes this model is considered. Characteristic function of the ergodic distribution of this process is expressed by means of the boundary functional of the random walk with a discrete interference of chance. Furthermore, under the assumption that the random variable �1 which describes the discrete interference of chance has a normal distribution with parameters (a;�2), an asymptotic expansion for the ergodic distribution of the stochastic process Wa(t) = (X(t) s)=a is obtained, as a ! 1. Moreover, it is proved that the ergodic distribution of the process Wa(t) weakly converges to the uniform distribution over the interval (0;1), as a ! 1. Finally, the accuracy of the approximation formula is tested by Monte-Carlo simulation method.