JAQM Volume 5, Issue 2 - June 30, 2010
Applications of Quantitative Methods to e-Commerce
A Simple Analysis of Customers Impatience in Multiserver Queues (p )
Balking and reneging are two ways through which customer impatience finds reflection. In the analysis of stochastic reneging, it has traditionally been assumed that the distribution of patience time is state independent. However in many queuing systems, the customer is aware of its position in the system state. Hence in this paper, we assume that a customer who arrives at the queuing system gets to know the state of the system. Consequently, both balking and reneging are taken as function of system state. Both types of reneging are considered. Explicit closed form expressions of a number of performance measures are presented. A numerical example is presented to demonstrate the results derived rounds off the paper.
Spam Filtering for Optimization in Internet Promotions using Bayesian Analysis (p )
The main characteristics of an e-business and its promoting are presented. It contains ways of promoting an e-business, examined in depth the e-mail marketing principle along with advantages and disadvantages of the implementation. E-mail marketing metrics are defined for analyzing the impact on customers. A model for optimization the promoting process via email is created for reaching the threshold of profitability for electronic business. The model implements Bayesian spam filtering and applies an internal classification, using the principles of conditioned probabilities.
WEB Services Integration with Distributed Applicationsn (p )
A Web service is a distributed application component. Web services distributed computing model allows application-to-application communication. There is nothing fundamentally new about the basic concept and the related technologies. The innovative thing about this is the reach of Web services and its ubiquitous support by literally all major vendors. Most likely, heterogeneity will at the end no longer be an obstruction for distributed applications. This paper describes the concept of service-oriented architecture (SOA) in conjunction with the Web services technology and covers the core Web services specifications which form a powerful and robust foundation for building distributed systems. It is presented a case study regarding the integration of the Web services with the SAP system for handling interoperability issues.
The conclusions and the future proposed developments are presented in the end of the paper.
E-Banking in Developing Economy: Empirical Evidence from Nigeria
Elisha Menson AUTA
This paper empirically examines the impact of e-banking in Nigeria’s economy using Kaiser-Meyar-Olkin (KMO) approach and Barlett’s Test of Sphericity which support the use of factor analysis in order to extract independent variables associated with e-banking. The paper explores the major factors responsible for internet banking based on respondents’ perception on various e-banking applications. It also provides a framework of the factors which are taken to assess the e-banking perception. Due to emergence of global economy, e-business has increasingly become a necessary component of business strategy and a strong catalyst for economic development. E-banking has become popular because of its convenience and flexibility, and also transaction related benefits like speed, efficiency, accessibility, etc. The results of this study shows that e-banking serves several advantages to Nigerian banking sector. The customers (respondents) perception is that e-banking provides convenience and flexible advantages. It also provides transaction related benefits like easy transfer, speedy transaction, less cost and time saving. However, the study shows that the Nigerian customers have security, access, and no enough knowledge regarding e-banking services rendering by banking sector in Nigeria. The study suggest that critical infrastructure like power and telecommunication should be provided and with high level of stability to ensure the application of e-banking in Nigeria. Also, the relative skewed nature of banks location mostly in urban area should be addressed to ensure spread and accessibility by rural dwellers.
A Study on Weibull Distribution for Estimating the Parameters (p )
The wind resource varies with of the day and the season of the year and even some extent from year to year. Wind energy has inherent variances and hence it has been expressed by distribution functions. In this paper, we present some methods for estimating Weibull parameters, namely, shape parameter ( k ) and scale parameter ( c ). The Weibul distribution is an important distribution especially for reliability and maintainability analysis. The suitable values for both shape parameter and scale parameters of Weibull distribution are important for selecting locations of installing wind turbine generators. The scale parameter of Weibull distribution also important to determine whether a wind farm is good or not. The presented method is the analytical methods and computational experiments on the presented methods are reported.
Minimization, Constraints and Composite Bézier Surfaces (p )
This paper presents a global method for approximation and/or construction of surfaces using constraints. The method is based on a min max problem which describes approximation and differential geometric characteristics, constrained in order to achieve desired geometrical effects. The numerical solution of the problem takes full advantage of the Finite-Elements method and of constrained optimization algorithms.
Comparative Multidimensional Analysis of Wheat Production at County Level (p )
This article presents a series of multidimensional analysis methods on agricultural factors at county level, performing subsequently a comparison between the results obtained in 2004 and, respectively, in 2008. By applying factor analysis, principal components analysis and hierarchical classes analysis, this article will seek to highlight the main variables that influence the agricultural production of grain, by territorial grouping by county, and then forming some specific county typologies for Romanian wheat production.
A Genetic Algorithm based Reliability Redundancy Optimization for Interval Valued Reliabilities of Components (p )
The goal of this paper is to solve the constrained redundancy allocation problem of series-parallel/parallel-series/complex system with the interval valued reliability of each component. For maximizing the overall system reliability under limited resource constraints, the problem is formulated as an unconstrained integer programming problem with interval coefficients by two different penalty techniques. To solve the transformed problem, we have developed two different real coded GAs for integer variables with tournament selection, uniform crossover, uniform mutation and different fitness functions based on penalty techniques. As a special case, considering the lower and upper bounds of the initial valued reliabilities of the component as the same, the corresponding problem has been solved. To illustrate the model, some numerical examples have been solved by our developed GAs and the results have been compared. Finally, to study the stability of our developed GAs with respect to the different GA parameters (like, population size, crossover and mutation rates), sensitivity analyses have been shown graphically.
Multinomial Logistic Regression: Usage and Application in Risk Analysis (p )
The objective of the article was to explore the usage of multinomial logistic regression (MLR) in risk analysis. In this regard, performing MLR on risk analysis data corrected for the non-linear nature of binary response and did address the violation of equal variance and normality assumptions. Additionally, use of maximum likelihood (-2log) estimation provided a means of working with binary response data. The relationship of independent and dependent variables was also addressed.
The data used included a cohort of hundred risk analyst of a historically black South African University. In this analysis, the findings revealed that the probability of the model chi-square (17.142) was 0.005, less than the level of significance of 0.05 (i.e. p<0.05). Suggesting that there was a statistically significant relationship between the independent variable-risk planning (Rp) and the dependent variable-control mechanism (control mecs) (p<0.05). Also, there was a statistically significant relationship between key risks assigned (KSA) and time spent on risk mitigation. For each unit increase in confidence in control mecs, the odds of being in the group of survey respondents who thought institution spend too little time on Rp decreased by 74.7%. Moreover, the findings revealed that survey respondents who had less confidence in control mecs were less likely to be in the group of survey respondents who thought institution spent about the right amount of time on risk planning.
A Mutivariate Strategic Perspective for the Evaluation of Customer Satisfaction in Great Distribution (p )
The author proposes a strategy for the analysis of data from Customer Satisfaction (CS) in Great Distribution. The aim of this paper is to evaluate CS through a comparison of multivariate statistical methodologies.
In this paper the author compares different estimations of Structural Equation Model (SEM) in a case study: evaluation of customer satisfaction in a supermarket. Overall satisfaction is determined by reference to three departments: “Salami and Cheese”, “Butchery”, “Fruit and Vegetables”. Each department is assessed through three aspects: “Assortment”, “Staff” and “Offer”.
Initially, the links between the different variables are verified through factor analysis and subsequently inserted into a structural equation model. To estimate the model the approach of “maximum likelihood” was used, with LISREL software. Finally, the “Partial Least Squares (PLS) approach was used to confirm the results.
The Coin Changing Problem as a Mathematical Model (p )
Mike C. PATTERSON,
The coin changing problem is a well-known problem to the general public and to operations research specialists and computer scientists. It is a popular topic for both discussion and programming assignments (Ghosh, 2008). The problem is a good example of the recursive nature of many problem solving techniques. Dynamic programming and greedy heuristics are perhaps the most common approaches to the coin changing problem (Johnsonbaugh and Schaefer, 2004).
Information Theoretic Estimation Improvement to the Nonlinear Gompertz’s Model Based on Ranked Set Sampling (p )
Amajd D. AL-NASSER
The aim of this paper is to apply both Generalized Maximum Entropy (GME) estimation method and Ranked Set Sampling (RSS) technique to improve the estimations of the Gompertz’s Model. The Gompertz’s model is a simple formula which expresses the geometrical relationship between the force of mortality and age.
The choice of evaluating the RSS is due to the fact that in many practical applications of the Gompertz’s model, as in biological or environmental sciences, the variable of interest is more costly to measure but is associated with several other easily obtainable.
In this paper, we have used Monte Carlo experiments to illustrate the performance of the GME estimator based on two different sampling techniques: the Simple Random Sample (SRS) and RSS. Moreover, the results are compared with the traditional Maximum Likelihood Estimates (MLE).
The Rasch Model for Evaluating Italian Student Performance (p )
In 1997 the Organisation for Economic Co-operation and Development (OECD) launched the OECD Programme for International Student Assessment (PISA) for collecting information about 15-year-old students in participating countries.
Our study analyse the PISA 2006 cognitive test for evaluating the Italian student performance in mathematics, reading and science comparing the results of different local governments. For this purpose the most proper statistic methodology is Item Response Theory - IRT that collects several models, the simplest is Rasch Model – MR (1960). As the items used in the analysis are both dichotomous that polytomous, we apply Partial Credit Model (PCM).
An Error Resilient and Memory Efficient Scheme for Wavelet Image Coding (p )
Irshad A. ARSHAD,
Set-Partitioning in Hierarchical Trees (SPIHT) is a state-of-art zero-tree based image coder with excellent rate-distortion performance in the noise free environments. However in presence of noise they are extremely sensitive to the bit errors. Even a single bit error may lead a substantial degradation in the quality. Our investigations reveal that the use of three linked lists in SPIHT increases the inter-bitplane dependencies of bits and therefore causes the errors to propagate down the bitstream. In this paper we propose a modified coder which uses only a single list. The list is initialized, processed and exhausted within each bit-plane, thereby removing dependencies of inter-plane bits. The proposed coder is known as Single List SPIHT (SL-SPIHT) image coder. Additionally, this coder is memory efficient as compared to the SPIHT. Simulation results show that under the same channel conditions, SL-SPIHT can improve the quality of reconstructed image by 5-6 dB compared to that of SPIHT.
A Cost Model for the IT Department (p )
Traditionally, it is believed that a busy IT Department (Information Technology Department) of an organization, with many services offered for other departments is very profitable, or at least “cost-effective”. A real and precise calculation of costs per product or service have unveiled that the above is not necessarily true. Activity-Based Costing is a costing model that could assign precise costs to products and services. The combination of this method with the model for calculating the total cost of an IT Department, proposed by HP Laboratories in [ ] is a powerful tool for improving IT services.
Book Review (p )
Sabry M. ABD-EL-FATTAH
Book Review on Structural Equation Modeling With Amos: Basic Concepts, Applications, and Programming (2nd Ed.)
by Barbara M. BYRNE