Cross validation (CV) is widely used for model assessment and comparison. In this thesis, we first review and compare three v-fold CV strategies: best single CV, repeated and averaged CV and double CV. The mean squared errors of the CV strategies in estimating the best predictive performance are illustrated by using simulated and real data examples. The results show that repeated and averaged CV is a good strategy and outperforms the other two CV strategies for finite samples in terms of the mean squared error in estimating prediction accuracy and the probability of choosing an optimal model. In practice, when we need to compare many models, conducting repeated and averaged CV strategy is not computational feasible. We develop an efficient sequential methodology for model comparison based on CV. It also takes into account the randomness in CV. The number of models is reduced via an adaptive, multiplicity-adjusted sequential algorithm, where poor performers are quickly eliminated. By exploiting matching of individual observations, it is sometimes even possible to establish the statistically significant inferiority of some models with just one execution of CV. This adaptive and computationally efficient methodology is demonstrated on a large cheminformatics data set from PubChem. Cross validated mean squared error (CVMSE) is widely used to estimate the prediction mean squared error (MSE) of statistical methods. For linear models, we show how CVMSE depends on the number of folds, v, used in cross validation, the number of observations, and the number of model parameters. We establish that the bias of CVMSE in estimating the true MSE decreases with v and increases with model complexity. In particular, the bias may be very substantial for models with many parameters relative to the number of observations, even if v is large. These results are used to correct CVMSE for its bias. We compare our proposed bias correction with that of Burman (1989), through simulated and real examples. We also illustrate that our method of correcting for the bias of CVMSE may change the results of model selection.

The performance of an electric utility vehicle is very dependent on how it is used. The range can decrease significantly if there are many uphill along the route and if the vehicle carry a heavy load. The most important is that the vehicle never stops because lack of energy or power. The purpose of this thesis was to study how performance of the electric utility vehicle Nimbell Trigo were affected, depending on route, drive pattern and load. To study the performance, the parameters: battery current, state of charge and speed were analyzed. To analyze these parameters, a dynamic model of the vehicle has been created with Simulink. The model is based on physical relations and vehicle data. With the specifications for the vehicle and the force acting on it, energy consumption, battery current and power demand have been calculated with the Simulink model for specific routes and drive patterns. To validate the mode, three different routes were driven with Nimbell Trigo in Gothenburg and a simulation of the same routes and drive patterns were made with the Simulink model. The result from the validation shows that the dynamic model can predict the parameters of interest well, particularly the state of charge. The Simulink model can in the future be a time and cost efficient way, to analyze the performance of Nimbell Trigo and thereby replace test drives in an initial phase. / Marknaden för elektriska arbetsfordon har ökat de senaste åren i takt med dels ökade krav på miljövänliga alternativ, dels framsteg inom batteritekniken. Akilleshälen för de elektriska fordonen som grupp har varit energilagringen i batterierna. Erfordrad batteri- och motorstorlek är starkt kopplat till användningen av fordonet. Vid körningar med långa uppförsbackar samt tung last, förkortas räckvidden markant. Vid för höga belastningar på fordonet minskar maximala hastigheten och i längden kan motorn överhettas. För användaren är det av stor vikt att fordonet inte blir stillastående för att effekten eller energin inte räcker till. Syftet med detta examensarbete var att utreda hur det elektriska arbetsfordonet Nimbell Trigo presterar för olika rutter och körmönster, med en given last. För att utreda prestandan av fordonet skapades därför en dynamisk modell av Nimbell Trigo i det dynamiska simuleringsverktyget Simulink. De parametrar som valdes att studera var batterinivå som är kopplat till energin, batteriströmmen som är relaterad till effekten och i sin tur påverkar batterinivån, samt hastigheten som kan indikera när effekten inte räcker till. Med den dynamiska modellen kan effektbehov och energikonsumtion för en given sträcka beräknas på förhand. Fordonets hastighet, acceleration samt backtagningsförmåga erhålls för olika situationer från simuleringen. Dessa parametrar kan därmed studeras med avseende på olika motoralternativ genom att ange moment- och verkningsgradskurvan för motorn. Den dynamiska modellen validerades med testkörningar där Nimbell Trigo kördes en viss rutt med ett givet körmönster och last. Motorfrekvens, batteriström och batterinivå uppmättes. Dessa parametrar simulerades i modellen för samma rutt och körmönster. Mätvärdena från testkörningarna plottades mot de värden som erhölls från simuleringen. Resultatet visar att den dynamiska modellen som skapades stämmer bra överens mot uppmätta testdata. Den dynamiska modellen kan i framtiden komma att användas, för att på ett kostnads- och tidseffektivt sätt ersätta testkörningar i initiala studier av olika drivlinor och batterier.

Kadangi įmonės atlieka naujo produkto įvedimą į rinką nesiremiant metodiškai pagrįsta, naujo produkto įvedimo į rinką matrica, naudoja tik atskirus privalomus tokios metodikos elementus, ir neturi tinkamų įvertinimo kriterijų naujo produkto įvedimo sėkmei pamatuoti pasitvirtino rekomenduotina naudoti darbo autoriaus pristatytą „Naujo produkto įvedimo į rinką modelį, kuris pagelbės įmonių konkurencingumo užtikrinimui. / The aim of the study is to formulate a model that Lithuanian enterprises could use in practice getting a new product entry into the market. The object of the study is the enterprises which are selling water heaters established in Kaunas city. The main argument of the study was that these enterprises do not have clearly delineated, prepared after methodological concepts new product entry to the market matrix, that the enterprises do not possess proper assessment criteria for the new product entry to the market success measurement. The author of the study has made a theoretical analysis of the papers and presented the new product entry to the market scientific approaches. It is estimated the new product entry to the market models proposed by the number of scientists in various papers and studies. The empirical research was held in 2005-2006 study years. 10 enterprises took part into the research. According to the research data and theory analysis the author formulated a new product entry to the market model that Lithuanian enterprises can use in their practice.

It is possible that the maximum likelihood method can give an inconsistent result when the DNA sequences are generated under a tree topology which is in the Felsentein Zone and analyzed with a misspeci ed model. Therefore, it is important to select a good substitution model. This thesis rst explores the e ects of di erent degrees and types of model misspeci cation on the maximum likelihood estimates. The results are presented for tree selection and branch length estimates based on simulated data sets. Next, two Pearson's goodness-of- t tests are developed based on binning of site patterns. These two tests are used for testing the adequacy of substitution models and their performances are studied on both simulated data sets and empirical data.

The proportional hazard (PH) cure rate model and the marginal structural Cox model (MSCM) are two broad areas used in analysing survival models with longitudinal data. Cure rate models were introduced to deal with survival models in the presence of a cure fraction and marginal structural models were introduced to adjust for time- ependent confounders through time-dependent weighting in longitudinal studies. However, few studies have tried to combine these two areas in building cure rate models in the presence of time-dependent covariates and time-dependent confounders. This thesis proposes an extension of the maximum likelihood estimation procedure for the PH cure rate model by incorporating (i) case weights, (ii) time-dependent covariates, and (iii) time-dependent weights in the presence of time-dependent covariates and time-dependent confounders into the model. Further, this thesis compares the performance of the PH cure rate model with case weights to the standard unweighted PH cure rate model through simulation studies. Results of these studies suggest that adding case weights in the PH cure rate model improves the estimation of the latency parameter when the sample size is relatively small.

The human hand has evolved to be remarkably good at skillfully manipulating objects. This manipulation requires knowledge of the dynamic properties of an object, which is represented in the central nervous system (CNS) by what has been referred to as an internal model. Internal models are neural representations of the predicted behaviour of objects or limbs with a known state in response to a given motor command. Our ability to successfully manipulate a wide variety of objects suggests that the CNS maintains multiple internal models of familiar object dynamics. People are able to both recruit these models for use when an object is grasped and to rapidly switch to another model when the object is exchanged. The purpose of this study was to investigate how internal models of objects are accessed and used for action. In experiment 1, subjects learned to move a cursor to a target by manipulating a robotic arm with complex dynamics. We used event-related fMRI to measure the neural activity associated with grasping the robot handle in preparation for movement. In comparison to control tasks, subjects showed significant neural activation in the ipsilateral cerebellum and the contralateral primary motor and supplementary motor areas, suggesting the likely involvement of these areas in recruitment of internal models. In experiment 2, we used a precision lifting task to investigate how the internal representation of weight asymmetry transfers across changes in hand, hand orientation and object orientation. Subjects demonstrated positive transfer in all cases when the hand was rotated, indicating that internal models of objects can be adapted to accommodate changes in hand orientation. When the object was rotated, positive transfer was seen only when the hand also rotated, suggesting that this change in hand orientation facilitated mental rotation of the object. Overall, these results support the idea that people maintain an internal representation of object dynamics but can not always link this model to the configuration of the object in space. / Thesis (Master, Neuroscience Studies) -- Queen's University, 2007-08-29 10:57:23.511

This paper will examine John Rawls’s views regarding the make up of society and, in particular, it will investigate the question of whether or not Rawls’s theory of justice is dependent upon accepting and endorsing the view that society is, or, at least, resembles an association. The aim of the paper will be to defend the position that Rawls’s theory of justice is in fact dependent upon an associational conception of society, and that it, therefore, relies upon the idea that the individuals in a society share a common interest or purpose. Thus, far from providing a liberal conception of society, this paper will attempt to illustrate that Rawls actually presents a more communitarian outlook of society than many might expect from a purportedly liberal thinker. In order to best prove the essential link to an associational conception of society, I discuss three principles that provided Rawls with reasons for preferring the difference principle to other modes of justice, and suggest that these founding principles provide sufficient evidence for connecting Rawls’s theory of justice to an associational conception of society. By advocating such a vision of society, Rawls betrays a fundamental premise of his own political liberalism, namely, the idea that individuals do not necessarily share a common interest or purpose with one another. / Thesis (Master, Philosophy) -- Queen's University, 2007-09-25 19:19:43.637

This thesis estimates the effects of monetary policy shocks by employing vector auto regressions (VAR). I argue that to the extent the central bank and the private sector have information not reflected in the VAR, the measurement of policy innovations is contaminated. These incorrectly estimated policy shocks then generate misleading results about the effects of monetary policy. This thesis first attempts to figure out the variables indeed observed by central banks to make monetary policy decisions and then formulates the monetary policy reaction function by using those variables. Having identified more realistic monetary policy functions in VAR models, I conclude that most of the previous puzzling results about the effect of monetary policy shocks might be due to incorrectly identifying the monetary policy reaction function. / Thesis (Ph.D, Economics) -- Queen's University, 2009-06-15 15:59:13.04

Knee osteoarthritis is a disease that affects nearly 40% of the global population over the age of seventy. It is believed that the incidence and progression of osteoarthritis can be partially attributed to changes in mechanical joint loading. Consequently, changes in lower-limb joint moments are important outcome measures for its treatment and prevention. The purpose of this study was to investigate the effect of four different anatomic joint models on the detection of changes in lower-limb joint moments due to knee osteoarthritis. Moments during gait were calculated for 44 subjects with moderate osteoarthritis and 44 asymptomatic subjects, then expressed using four joint models: Joint Coordinate System, Plane of Progression, Distal, and Proximal. Discrete peak measures and principal component scores were compared between groups. Hip adduction magnitude, knee adduction magnitude, peak early-stance knee internal rotation, and peak ankle plantarflexion moments were different between groups regardless of joint model. Differences detected using principal component analysis were less sensitive to the choice of joint model. Results support adoption of the Joint Coordinate System as a standard for joint moment expression due to its clinical relevance and ability to detect differences due to moderate knee osteoarthritis. / Thesis (Master, Mechanical and Materials Engineering) -- Queen's University, 2009-07-29 14:08:56.683

The objective of this work is to develop mathematical models for predicting the thermal stability of commercial diagnostic assays. These assays are a product of the Point of Care division of Abbott laboratories, and are used for analyzing patient blood samples for specific substances. The accuracy of the results from these diagnostic tests relies on the activity of specific biological and/or chemical components of the sensors. Mathematical models that describe the stability of these active components are useful for supporting product shelf-life claims and for the design and implementation of accelerated testing protocols. In the thesis, the stability of two diagnostic assay systems of interest to Abbott Point of Care is investigated using mathematical modeling. For the first assay system investigated, the biosensor associated with the assay is identified as an important factor for product stability. A second-order dynamic model is developed to describe the thermal stability of this biosensor. The model corresponds to a reversible reaction followed by an irreversible reaction, with rate coefficients having Arrhenius temperature dependencies. The second-order dynamic model provides improved predictions relative to a first-order dynamic model, based on a comparison between model fits for two experimental datasets, and a comparison of predictive ability for a validation dataset. The second-order dynamic model is used to extend the concept of Mean Kinetic Temperature concept from the pharmaceutical industry to systems with higher-order dynamics. For the second assay system investigated, the calibration fluid is identified as a key factor in assay stability. A first-order model is developed to describe the stability of the analyte within the calibration fluid. The first-order model captures most of the trend present in the data from calibration fluid incubation experiments. Finally, model predictions are used to investigate the amount of change in assay response that can be attributed to changes in concentration of analyte in the calibration fluid (after storage at elevated temperatures). The results show that the changes observed in assay responses are consistent with the magnitude of changes in calibrant analyte concentrations predicted by the model. / Thesis (Master, Chemical Engineering) -- Queen's University, 2011-02-02 00:09:23.758