Sami M. Halawani

A rule-based method for identifying the factor structure in customer satisfaction

The analysis of customer satisfaction datasets has shown that product-related features fall into three categories (i.e., basic, performance, and excitement), which affect overall satisfaction differently. Because the relationship between product features and customer satisfaction is characterized by non-linearity and asymmetry, feature values are studied to understand the characteristics of a feature. However, existing methods are computationally expensive and work for ordinal features only. We propose a rule-based method that can be used to analyze data features regarding various characteristics of customer satisfaction. The inputs for these rules are derived by using a probabilistic feature-selection technique. In this feature selection method, mutual associations between feature values and class decisions in a pre-classified database are computed to measure the significance of feature values. The proposed method can be used for both types of features: ordinal and categorical. The proposed method is more computationally efficient than previously recommended methods. We performed experiments on a synthetic dataset with known characteristics, and our method correctly predicted the characteristics of the dataset. We also performed experiments with a real-housing dataset. The knowledge extracted from the dataset by using this method is in agreement with the domain knowledge.

Improvement of decision making grid model for maintenance management in Small and Medium Industries

Small and Medium Industries (SMIs) generate significant contribution in the economic growth in developing countries. To achieve the standard expected product from customers, SMIs must produce marketable product in term of quality, quantity and prices. This requires the investment in technology for supporting daily operation and maintenance management. In previous research, maintenance decision support system has been developed using decision making grid (DMG) model to solve the problem. In this study, the researches apply tri-quadrant technique to cluster the DMG. This study focuses on the effect from the improvement of DMG model in categorizing low, medium and high criterion for the real production floor in SMIs. The result shows not only the reduction of cost and machines downtime but also better reliability on daily maintenance operation.

A versatile mathematical work-flow to explore how Cancer Stem Cell fate influences tumor progression

BackgroundNowadays multidisciplinary approaches combining mathematical models with experimental assays are becoming relevant for the study of biological systems. Indeed, in cancer research multidisciplinary approaches are successfully used to understand the crucial aspects implicated in tumor growth. In particular, the Cancer Stem Cell (CSC) biology represents an area particularly suited to be studied through multidisciplinary approaches, and modeling has significantly contributed to pinpoint the crucial aspects implicated in this theory.More generally, to acquire new insights on a biological system it is necessary to have an accurate description of the phenomenon, such that making accurate predictions on its future behaviors becomes more likely. In this context, the identification of the parameters influencing model dynamics can be advantageous to increase model accuracy and to provide hints in designing wet experiments. Different techniques, ranging from statistical methods to analytical studies, have been developed. Their applications depend on case-specific aspects, such as the availability and quality of experimental data, and the dimension of the parameter space.ResultsThe study of a new model on the CSC-based tumor progression has been the motivation to design a new work-flow that helps to characterize possible system dynamics and to identify those parameters influencing such behaviors. In detail, we extended our recent model on CSC-dynamics creating a new system capable of describing tumor growth during the different stages of cancer progression. Indeed, tumor cells appear to progress through lineage stages like those of normal tissues, being their division auto-regulated by internal feedback mechanisms. These new features have introduced some non-linearities in the model, making it more difficult to be studied by solely analytical techniques. Our new work-flow, based on statistical methods, was used to identify the parameters which influence the tumor growth. The effectiveness of the presented work-flow was firstly verified on two well known models and then applied to investigate our extended CSC model.ConclusionsWe propose a new work-flow to study in a practical and informative way complex systems, allowing an easy identification, interpretation, and visualization of the key model parameters. Our methodology is useful to investigate possible model behaviors and to establish factors driving model dynamics.Analyzing our new CSC model guided by the proposed work-flow, we found that the deregulation of CSC asymmetric proliferation contributes to cancer initiation, in accordance with several experimental evidences. Specifically, model results indicated that the probability of CSC symmetric proliferation is responsible of a switching-like behavior which discriminates between tumorigenesis and unsustainable tumor growth.

In silico characterization of a putative ORF-MAP1138c of Mycobacterium avium subspecies paratuberculosis (MAP) with its implications in virulence

Background Johne’s disease is a chronic mycobacterial infection of the small intestine affecting ruminants worldwide. It is estimated that over 50% of the European Union (EU) dairy holdings is infected [1]. The causal agent is Mycobacterium avium subspecies paratuberculosis (MAP), a slow-growing, acid-fast bacterium. It is a part of the Mycobacterium avium complex (MAC), which also comprises of opportunistic pathogens of humans, as well as innocuous, environmental bacteria [2]. MAP generally interacts with macrophages via different types of receptors, including Toll-like receptors (TLRs) [3,4]. It has been demonstrated of late that H37Rv1411c (LprG) enhances the recognition of triacylated Mycobacterium tuberculosis glycolipids by TLR2 and thereby restraining the expression of MHC-II molecules and processing of antigen and presentation of MHC restricted antigens by macrophages in a TLR2dependent manner [5,6]. However, little is known about how M. paratuberculosis evades and resists this active CD4 T-cell response and survives and infects other macrophages, a hallmark of mycobacterial infections. In this context, the identification of antigenic proteins is useful in understanding the immune evasion mechanism of MAP within host macrophages.

Novel ensemble methods for regression via classification problems

Regression via classification (RvC) is a method in which a regression problem is converted into a classification problem. A discretization process is used to covert continuous target value to classes. The discretized data can be used with classifiers as a classification problem. In this paper, we use a discretization method, Extreme Randomized Discretization (ERD), in which bin boundaries are created randomly to create ensembles. We present two ensemble methods for RvC problems. We show theoretically that the proposed ensembles for RvC perform better than RvC with the equal-width discretization method. We also show the superiority of the proposed ensemble methods experimentally. Experimental results suggest that the proposed ensembles perform competitively to the method developed specifically for regression problems.

A Costing Analysis for Decision Making Grid Model in Failure-Based Maintenance

Background. In current economic downturn, industries have to set good control on production cost, to maintain their profit margin. Maintenance department as an imperative unit in industries should attain all maintenance data, process information instantaneously, and subsequently transform it into a useful decision. Then act on the alternative to reduce production cost. Decision Making Grid model is used to identify strategies for maintenance decision. However, the model has limitation as it consider two factors only, that is, downtime and frequency of failures. We consider third factor, cost, in this study for failure-based maintenance. The objective of this paper is to introduce the formulae to estimate maintenance cost. Methods. Fish bone analysis conducted with Ishikawa model and Decision Making Grid methods are used in this study to reveal some underlying risk factors that delay failure-based maintenance. The goal of the study is to estimate the risk factor that is, repair cost to fit in the Decision Making Grid model. Decision Making grid model consider two variables, frequency of failure and downtime in the analysis. This paper introduces third variable, repair cost for Decision Making Grid model. This approaches give better result to categorize the machines, reduce cost, and boost the earning for the manufacturing plant. Results. We collected data from one of the food processing factories in Malaysia. From our empirical result, Machine C, Machine D, Machine F, and Machine I must be in the Decision Making Grid model even though their frequency of failures and downtime are less than Machine B and Machine N, based on the costing analysis. The case study and experimental results show that the cost analysis in Decision Making Grid model gives more promising strategies in failure-based maintenance. Conclusions. The improvement of Decision Making Grid model for decision analysis with costing analysis is our contribution in this paper for computerized maintenance management system.

Interaction between Sunlight and the Sky Colour with 3D Objects in the Outdoor Virtual Environment

The sky has always been the crucial element in modeling the background of an outdoor scene. The position of the sun during the day gives a different impact on the sky colour. The sky colour indirectly affects the colour of the objects which were exposed to the lighting, such as the orangish red colour of the clouds seen during sunsets. Consequently, this study will emphasize on how to produce illuminated 3D objects based upon the effects of interaction between the sunlight and sky. A two-part program was developed for this study. The first part of the program concentrates on producing the correct sky colour depending on the position of the sun using Perez’s function. The sky colour will be plotted on the sky dome which in turn will become a closed environment for the clouds. The interaction will occur in the second part of the program where the energy transfer in the dome environment with color of the sky as the main source illumination, resulting in the colour bleeding effect when using the radiosity approach. The result from this study is applicable to daylight modeling of building by showing the lighting effects from the sun and the sky.

Consistency of randomized and finite sized decision tree ensembles

Regression via classification (RvC) is a method in which a regression problem is converted into a classification problem. A discretization process is used to covert continuous target value to classes. The discretized data can be used with classifiers as a classification problem. In this paper, we use a discretization method, Extreme Randomized Discretization, in which bin boundaries are created randomly to create ensembles. We present an ensemble method for RvC problems. We show theoretically for a set of problems that if the number of bins is three, the proposed ensembles for RvC perform better than RvC with the equal-width discretization method. We use these results to show that infinite-sized ensembles, consisting of finite-sized decision trees, created by a pure randomized method (split points are created randomly), are not consistent. We also theoretically show, using a set of regression problems, that the performance of these ensembles is dependent on the size of member decision trees.

Ensemble methods for prediction of parkinson disease

Parkinson disease is a degenerative disorder of the central nervous system. In the present paper, we study the effectiveness of regression tree ensembles to predict the presence and severity of symptoms from speech datasets. This is a regression problem. Regression via classification (RvC) is a method in which a regression problem is converted into a classification problem. A discretization process is used to convert continuous target value to classes. The discretized data can be used with classifiers as a classification problem. In this paper, we also study a recently developed RvC ensemble method for the prediction of Parkinson disease. Experimental results suggest that the RvC ensembles perform better than a single regression tree. Experiments also suggest that regression tree ensembles created using bagging procedure can be a useful tool for predicting Parkinson disease. The RvC ensembles and regression tree ensembles performed similarly on the dataset.

Use of Flow Equivalent Servers in the Transient Analysis of Product Form Queuing Networks

In this paper we deal with approximate transient analysis of Product Form Queuing Networks. In particular, we exploit the idea of flow equivalence to reduce the size of the model. It is well-known that flow equivalent servers lead to exact steady state solution in many cases. Our goal is to investigate the applicability of flow equivalence to transient analysis. We show that exact results can be obtained even in the transient phase, but the definition of the equivalent server requires the analysis of the whole original network. We propose thus to use approximate aggregate servers whose characterization demands much less computation. Specifically, the characterization corresponds to the steady state equivalent server of the stations that we aim to aggregate and thus can be achieved by analyzing the involved stations in isolation. This way, approximations can be derived for any queuing network, but the precision of the results depends heavily on the topology and on the parameters of the model. We illustrate the approach on numerical examples and briefly discuss a set of criteria to identify the cases when it leads to satisfactory approximation.