Search

This paper proposes a genetic algorithm-based Markov Chain approach that can be used for non-parametric estimation of regression coefficients and their statistical confidence bounds. The proposed approach can generate samples from an unknown probability density function if a formal functional form of its likelihood is known. The approach is tested in the non-parametric estimation of regression coefficients, where the least-square minimizing function is considered the maximum likelihood of a multivariate distribution. This approach has an advantage over traditional Markov Chain Monte Carlo methods because it is proven to converge and generate unbiased samples computationally efficiently. [ABSTRACT FROM AUTHOR]

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ejor.2005.07.029 Byline: Parag C. Pendharkar, Girish H. Subramanian Keywords: Software engineering; Learning curve theory; Probability bounds; Chernoff bound; Neural networks; Central limit theorem; Moment generating functions Abstract: In this paper, we investigate the existence of learning curves in software development. Under the assumption of independent and identical distribution (iid) of programmer's experience and identical effort-experience learning curve relationship for different programmers, we illustrate the existence of an exponentially decreasing learning curve relationship between a programmer's effort and his/her ICASE tool experience, and show that the effort-experience relationship is inelastic when a programmer's ICASE tool experience is low. We analyze the impact of our assumptions on actual software development effort, and propose a tight probability upper bound and a central-limit theorem based probability estimator for estimating the approximate probability that the software development effort will be less than or equal to a certain number. Examples to illustrate the use of the probability estimator are also provided. Author Affiliation: School of Business Administration, Penn State Harrisburg, 777 West Harrisburg Pike, Middletown, PA 17078, United States Article History: Received 15 October 2004; Accepted 15 July 2005

A major step in effectively managing radio resources in a cellular network is to design an appropriate scheme for assigning cells to a location area (LA), serviced by a switch, and allocate resources for individual switches. However, this assignment is already proven in the literature to be an NP-hard problem [Merchant A, Sengupta B. Assignment of cells to switches in PCS networks. IEEE/ACM Transactions on Networking 3(5) (1995) 521-6] that requires efficient heuristic search techniques for obtaining real-time solutions. This work presents a state-space search technique, which is a variant of best first search heuristic technique. The algorithm called the block depth first search (BDFS), allocates cells to switches during switch level resource planning. Under various simulated performance criteria, we compare the performance of the proposed technique with other similar procedures in the literature. Our results indicate that the BDFS outperforms the meta-heuristic procedures in terms of both efficiency and quality of solutions. Hence, we conclude that our proposed technique can be effectively used for switch level planning leading to an efficient management of scarce radio resource in cellular networks. Keywords: Cell to switch assignment; Radio resource management; Depth first search, 1. Introduction In order to utilize better the scarce radio spectrum in mobile communication systems, all available radio resources (transmitter powers, channels and base stations) must be used in the [...]

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ejor.2005.10.050 Byline: Parag C. Pendharkar (a), Gary J. Koehler (b) Keywords: Genetic algorithms; Heuristic search; Worst-case analysis; Error bounds; Markov processes Abstract: We propose two general stopping criteria for finite length, simple genetic algorithms based on steady state distributions, and empirically investigate the impact of mutation rate, string length, crossover rate and population size on their convergence. Our first stopping criterion is based on the second largest eigenvalue of the genetic algorithm transition matrix, and the second stopping criterion is based on minorization conditions. Author Affiliation: (a) Information Systems, School of Business Administration, Capital College, Pennsylvania State University, 777 West Harrisburg Pike, Middletown, PA 17057, United States (b) Decision and Information Sciences, 351 BUS, The Warrington College of Business Administration, University of Florida, Gainesville, FL 32611, United States Article History: Received 27 July 2004; Accepted 7 October 2005

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ejor.2004.10.023 Byline: Parag C. Pendharkar Keywords: Data envelopment analysis; Artificial neural networks; Production function estimation Abstract: In this study, we investigate the factors that influence the object-oriented (OO) component size and source code documentation. For multiple inputs and multiple outputs, we use data envelopment analysis to illustrate that non-linear variable returns to scale (VRS) economies exist for OO component size and source code documentation. The existence of non-linear variable returns to scale economies indicates that non-linear regression models will perform better than linear regression models. Using empirical data, we compare the performance of non-linear artificial neural network (ANN) forecasting model and linear regression model. Our results indicate that the ANN model performs well when VRS economies exist between multiple inputs and multiple outputs. Author Affiliation: School of Business Administration, Penn State Harrisburg, 777 West Harrisburg Pike, Middletown, PA 17057, USA Article History: Received 27 January 2004; Accepted 30 October 2004

In this paper, we show how the data envelopment analysis (DEA) model might be useful to screen training data so a subset of examples that satisfy monotonicity property can be identified. Using real-world health care and software engineering data, managerial monotonicity assumption, and artificial neural network (ANN) as a forecasting model, we illustrate that DEA-based data screening of training data improves forecasting accuracy of an ANN. Index Terms--Data envelopment analysis, artificial neural networks, data preprocessing.

Many information systems and computer-mediated communication studies measure the users perceptions of email. The results indicate that an individualEs perceptions of email are a multidimensional construct with two dimensions that are the individual level and the organizational level.

Component‐based software development (CBSD) is an active area of research. Ascertaining the quality of components is important for overall software quality assurance in CBSD. One of the important metrics for measuring defects, analyzability, efforts, and cost in CBSD is component size. The paper presents an analytical model based on maximization of Tsallis entropy to obtain closed form expression for component size distribution (maximum Tsallis entropy component size distribution, MTECSD) in steady state. It is found that the component size distribution follows power law asymptotically. A procedure based on generalized Jensen–Shannon measure is developed to estimate model parameters. A detailed analysis of many popular probability distributions along with MTECSD is carried out on many diverse real data sets of component‐based softwares. The analysis reveals that lognormal and MTECSD distributions fit well to component sizes in many software conforming the presence of power law behavior. The software whose component size distributions are described by MTECSD are in equilibrium implying that new defects in these software systems occur occasionally. Power law behavior in component sizes also imply high variation leading to difficulty in software analyzability. The precise knowledge of component size distribution also provides an alternative method to compute efforts and cost estimates by modified COCOMO model. [ABSTRACT FROM AUTHOR]