This research assesses the performance of over-dispersed Poisson regression model and negative binomial model with count data. It examines the association between price plan features of mobile phone services and the number of people who adopt the plan. Mobile service data is used to estimate the model with a sample of one million customers running from February 2006 to September 2009. Under three main categories, customer type, age, and handset price, we run the model based on price plan features. Estimates are derived from the maximum likelihood estimation (MLE) method. Root mean squared error (RMSE) is used to observe the statistical fits of all the regression models. Then, we construct four estimation and holdout samples, leaving out one, three, six, and twelve months. The estimation constitutes the in-sample (IS) and the holdout represents the out-sample (OS). By estimating the IS, we predict the OS. Root mean squared error of prediction (RMSEP) is checked to see how accurate the prediction is. Results generally suggest that academic year start (AYS), seasonality, duration of months since launch of price plan (DMLP), basic fees, rate with no discount (RND), free call minutes (FCM), free data (FD), free text messaging (FTM), free perk rating (FPR), and handset support all show significant effect. The significance occurs depending on the segment. The RMSE and RMSEP show that the over-dispersed Poisson model outperforms the negative binomial model. Further implications and limitations of the results are discussed. / text

The primary controls on subsurface reservoir heterogeneities and fluid flow characteristics are sedimentary facies architecture and petrophysical rock fabric distribution in clastic reservoirs and in carbonate reservoirs, respectively. Facies models are critical and fundamental for summarizing facies and facies architecture in data-rich areas. Facies models also assist in predicting the spatial architectural trend of sedimentary facies in other areas where subsurface information is lacking. The method for transferring geological information from different facies models into digital data and then generating associated numerical models is called facies modeling or geological modeling. Facies modeling is also vital to reservoir simulation and reservoir characterization analysis. By extensively studying and reviewing the relevant research in the published literature, this report identifies and analyzes the best and most detailed geologic data that can be used in facies modeling, and the most current geostatistical and stochastic methods applicable to facies modeling. Through intensive study of recent literature, the author (1) summarizes the basic concepts and their applications to facies and facies models, and discusses a variety of numerical modeling methods, including geostatistics and stochastic facies modeling, such as variogram-based geostatistics modeling, object-based stochastic modeling, and multiple-point geostatistics modeling; and (2) recognizes that the most effective way to characterize reservoir is to integrate data from multiple sources, such as well data, outcrop data, modern analogs, and seismic interpretation. Detailed and more accurate parameters using in facies modeling, including grain size, grain type, grain sorting, sedimentary structures, and diagenesis, are gained through this multidisciplinary analysis. The report concludes that facies and facies models are scale dependent, and that attention should be paid to scale-related issues in order to choose appropriate methods and parameters to meet facies modeling requirements. / text

In order to explore a quantum version of a discrete nonlinear Schrödinger equation (DNLS), we quantize one nonlinear Schrödinger model, which is used to study different physical systems, e.g. coupled Bose-Einstein condensates. We will focus on small systems, like Dimer and Trimer.In our efforts to solve this quantum problem, we develop a Mathematica routine that implements the Number State Method and solves the corresponding Schrödinger equation. We calculate analytically and numerically the energy spectrum of the Dimer and Trimer systems. Those eigenenergies depend on the parameter set Q=Q1, Q2, Q3, Q4, Q5 and by adjusting this set Q, we can obtain the desired results and examine their effects. After the quantization of the extended DNLS we obtain a quantum DNLS, also known as an extended Bose-Hubbard (BH) model. The aim of this Master's thesis is to study the differences and similarities between the classical DNLS and the extended BH model, and what happens when we approach from the quantum regime to the classical one. Taking into account that the Hamiltonian has an important conserved quantity, the number operator, enables the total Hamiltonian to be block-diagonalized. This can be accomplished by taking advantage of additional symmetries, such as translational symmetry, which will simplify the analysis of the Hamiltonian matrix. In our results we discuss several effects that break the lattice symmetry, as the intersection between symmetric and antisymmetric states. We also compare our results with those obtained in previous works for the classical model, and we find some similarities, e.g. the transition of the highest-energy state from a one-site solution to a two-site solution depending on which Q parameters we vary, but also differences, as the appearance of a three-site solution, in a Trimer system.

In recent years, smartphone has become the most popular products, the literature for smartphone is relatively less, most of using Technology Acceptance Model ¡]TAM¡^ as the main research framework to explore. Influencing the consumer intension is concept of cognition ¡]perceived usefulness, perceived ease of use¡^, thereby affecting subsequent behavioral intentions, but what kind of product characteristics affect consumer¡¦s perceiving are seldom addressed. This study proposes "Involvement - Brand Loyalty Model" as the theoretical basis, with the "involvement" concept applied to smartphone users in order to explore the brand loyalty. Involvement antecedent focuses on "product utility", and explores what kind of product characteristics affect the product utility from the past literature, and then is combined with the social influence to investigate smartphone usage. Finally, the study will also compare different groups based on with or without owning a smartphone, and then provide the final analysis of the research results. The final results show that user involvement does affect the brand loyalty of the smartphone, and if the users are highly involved, regardless of whatever the user experience may be, product involvement will affect the brand involvement. Comparison with previous studies finds the similarity and the differences between those two groups. The common is the high involvement of smartphone. The differences lie in the user experience that does affect the user for the demand of products characteristics, and the impact of the social influence. The three products characteristics: convergence, innovation and network externality will definitely affect the product utility for someone who has smartphone, but we can understand that the user experience will impact convergence by analyzing the groups without smartphone. It is difficult to measure the value of convergence for someone who doesn¡¦t have smartphone, so in the study show that the convergence does not significantly affect the product utility, while the other two product characteristics ¡]innovation, network externalities¡^ are significant impact the product utility. For people who have usage experience, the social influence can directly affect their brand judgment, nevertheless without the usage experience, the social influence has impact on the brand involvement only through the product involvement.

It has been known that volatile oil and gas condensate reservoirs cannot be modeled accurately with conventional black-oil models. One variation to the black-oil approach is the modified black-oil (MBO) model that allows the use of a simple, and less expensive computational algorithm than a fully compositional model that can result in significant timesaving in full field studies. The MBO model was tested against the fully compositional model and performances of both models were compared using various production and injection scenarios for a rich gas condensate reservoir. The software used to perform the compositional and MBO runs were Eclipse 300 and Eclipse 100 versions 2002A. The effects of black-oil PVT table generation methods, uniform composition and compositional gradient with depth, initialization methods, location of the completions, production and injection rates, kv/kh ratios on the performance of the MBO model were investigated. Vertical wells and horizontal wells with different drain hole lengths were used. Contrary to the common belief that oil-gas ratio versus depth initialization gives better representation of original fluids in place, initializations with saturation pressure versus depth gave closer original fluids in place considering the true initial fluids in place are given by the fully compositional model initialized with compositional gradient. Compared to the compositional model, results showed that initially there was a discrepancy in saturation pressures with depth in the MBO model whether it was initialized with solution gas-oil ratio (GOR) and oil-gas ratio (OGR) or dew point pressure versus depth tables. In the MBO model this discrepancy resulted in earlier condensation and lower oil production rates than compositional model at the beginning of the simulation. Unrealistic vaporization in the MBO model was encountered in both natural depletion and cycling cases. Oil saturation profiles illustrated the differences in condensate saturation distribution for the near wellbore area and the entire reservoir even though the production performance of the models was in good agreement. The MBO model representation of compositional phenomena for a gas condensate reservoir proved to be successful in the following cases: full pressure maintenance, reduced vertical communication, vertical well with upper completions, and producer set as a horizontal well.

The intricate dependency of networking protocols upon the performance of the wireless channel motivates the investigation of network channel approximations for fading channels. Wireless networking protocols are increasingly being designed and evaluated with the assistance of networking simulators. While evaluating networking protocols such as medium access control, routing, and reliable transport, the network channel model, and its associated capacity, will drastically impact the achievable network throughput. Researcher relying upon simulation results must therefore use extreme caution to ensure the use of similar channel models when performing protocol comparisons. Some channel approximations have been created to mimic the behavior of a fading environment, however there exists little to no justification for these channel approximations. This dissertation addresses the need for a computationally efficient fading channel approximation for use in network simulations. A rigorous flat fading channel model was developed for use in accuracy measurements of channel approximations. The popular two-state Markov model channel approximation is analyzed and shown to perform poorly for low to moderate signal-to-noise ratios (SNR). Three novel channel approximations are derived, with multiple methods of parameter estimation. Each model is analyzed for both statistical performance and network performance. The final model is shown to achieve very accurate network throughput performance by achieving a very close matching of the frame run distributions. This work provides a rigorous evaluation of the popular two-state Markov model, and three novel low complexity channel models in both statistical accuracy and network throughput performance. The novel models are formed through attempts to match key statistical parameters of frame error run and good frame run statistics. It is shown that only matching key parameters is insufficient to achieve an acceptable channel approximation and that it is necessary to approximate the distribution of frame error duration and good frame run duration. The final novel channel approximation, the three-state run-length model, is shown to achieve a good approximation of the desired distributions when some key statistical parameters are matched.

Technological and methodological advances of the past few decades have provided hydrologists with advanced and increasingly complex hydrological models. These models improve our ability to simulate hydrological systems, but they also require a lot of detailed input data and, therefore, have a limited applicability in locations with poor data availability. From a case study of Big Creek watershed, a 186.4 km2 urbanizing watershed in Atlanta, GA, for which continuous flow data are available since 1960, this project investigates the relationship between model complexity, data availability and predictive performance in order to provide reliability factors for the use of reduced complexity models in areas with limited data availability, such as small ungaged watersheds in similar environments. My hope is to identify ways to increase model efficiency without sacrificing significant model reliability that will be transferable to ungaged watersheds.

Change in adaptive outcomes of children who are treated for brain tumors is examined using longitudinal data. The children received different types of treatment from none to any combinations of three treatments, which are surgery, radiation and chemotherapy. In this thesis, we use mixed model to find the significant variables that predict change in outcomes of communication skill, daily living skills and socialization skill. Fractional polynomial transformation method and Gompertz method are applied to build non-linear longitudinal curves. We use PRESS as the criterion to compare these two methods. Comparison analysis shows the effect of each significant variable on adaptive behaviors over time. In most cases, model with Gompertz method is better than that with Transformation method. Significant predictors of change in adaptive outcomes include Time, Gender, Surgery, SES classes, interaction between Time and Radiation, interaction between Time and Gender, interaction between Age and Gender.

We perform a review of model selection procedures, in particular various cross validation procedures and adaptive model selection. We cover important results for these procedures and explore the connections between different procedures and information criteria.

No description available.