Thesis advisor: Susanto Basu / This dissertation consists of three essays that study macroeconomic modeling and its application with a particular focus on monetary economics. In Chapter 1, I develop a New Keynesian model with heterogeneous workers whose wage settings are subject to downward nominal wage rigidity (DNWR) to address two puzzles of inflation dynamics: the missing deflation during the Great Recession and the excessive disinflation afterward. I demonstrate that DNWR introduces a time-varying wedge between the output gap and the marginal cost of producing one unit of output, which makes the observed Phillips curve flatter during recessions. Endogenous evolution of cross-sectional wage distribution generates various dimensions of non-linearities, while the presence of the zero lower bound (ZLB) of the nominal interest rate further reinforces the mechanism. Consequently, the model can quantitatively account for the inflation dynamics during and after the Great Recession under plausible parameter values that are consistent with micro evidence. In Chapter 2, I study welfare-maximizing monetary policy rule in the heterogeneous agent New Keynesian model with DNWR that is developed in Chapter 1. The optimal monetary policy rule responds strongly to output to address the inefficiency generated by DNWR, while responsiveness to inflation plays a minor role in welfare. Moreover, monetary policy can improve social welfare by responding more aggressively to a contractionary shock than to an expansionary one to offset the asymmetry stemming from DNWR. In the presence of the ZLB, on the other hand, alternative policy rules such as forward guidance and price-level targeting can partly offset the adverse effects of it by committing to a future low interest rate policy. I also investigate the optimal steady-state inflation rate. In Chapter 3, which is coauthored with Dongho Song and Jenny Tang, we propose a method of introducing theory-driven priors into the estimation of the vector autoregression (VAR). Our methodology is more flexible than existing methods in that it allows a researcher to incorporate prior beliefs on a subset of variables in theoretical models that are of key interest while remaining agnostic about other variables in the VAR. We apply to the problem of exchange rate forecasting for the British pound versus the US dollar. By imposing different combinations of priors informed by uncovered interest rate or purchasing power parity, we find that substantial gains are realized at longer forecast horizons. / Thesis (PhD) — Boston College, 2018. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Economics.

The electrical activity of skeletal muscle—the electromyogram (EMG)—is of value to many different application areas, including ergonomics, clinical biomechanics and prosthesis control. For many applications, the EMG is related to muscular tension, joint torque and/or applied forces. In these cases, a goal is for an EMG-torque model to emulate the natural relationship between the central nervous system (as evidenced in the surface EMG) and peripheral joints and muscles. This thesis work concentrated on experimental investigations of EMG-torque modeling. My contributions include: 1) continuing to evaluate the advantage of advanced EMG amplitude estimators, 2) studying system identification techniques (regularizing the least squares fit and increasing training data duration) to improve EMG-torque model performance, and 3) investigating the influence of joint angle on EMG-torque modeling. Results show that the advanced EMG amplitude estimator reduced the model error by 21%—71% compared to conventional estimators. Use of the regularized least squares fit with 52 seconds of training data reduced the model error by 20% compared to the least squares fit without regulation when using 26 seconds of training data. It is also demonstrated that the influence of joint angle can be modeled as a multiplicative factor in slowly force-varying and force-varying contractions at various, fixed angles. The performance of the models that account for the joint angle are not statistically different from a model that was trained at each angle separately and thus does not interpolate across angles. The EMG-torque models that account for joint angle and utilize advanced EMG amplitude estimation and system identification techniques achieved an error of 4.06±1.2% MVCF90 (i.e., error referenced to maximum voluntary contraction at 90° flexion), while models without using these advanced techniques and only accounting for a joint angle of 90° generated an error of 19.15±11.2% MVCF90. This thesis also summarizes other collaborative research contributions performed as part of this thesis. (1) EMG-force modeling at the finger tips was studied with the purpose of assessing the ability to determine two or more independent, continuous degrees of freedom of control from the muscles of the forearm [with WPI and Sherbrooke University]. (2) Investigation of EMG bandwidth requirements for whitening for real-time applications of EMG whitening techniques [with WPI colleagues]. (3) Investigation of the ability of surface EMG to estimate joint torque at future times [with WPI colleagues]. (4) Decomposition of needle EMG data was performed as part of a study to characterize motor unit behavior in patients with amyotrophic lateral sclerosis (ALS) [with Spaulding Rehabilitation Hospital, Boston, MA].

A mathematical modeling framework for the methane steam reforming reaction operating in steady state has been developed. Performances are compared between the classic catalytic packed bed reactor and a Pd-based catalytic membrane reactor. Isothermal simulations on MATLAB © has first been conducted and show a higher performance of the membrane reactor over the packed bed reactor. Methane conversion of 1 can be reached for lower temperatures than used with industrial PBR, and better performances are shown for an increase in the operating pressure. Optimum conditions were defined for Temperature (500-600 Celsius), reaction side pressure (16-40 bars), membrane thickness (1-7 micrometers), steam/methane ratio (3-4), reactor length (5-10 meters) and permeate sweep ratio (20 or more). This model was validated by multiple recognized sources. Adiabatic simulations were conducted in order to develop a mathematical model base for non-isothermal simulations. The membrane reactor is again showing a higher conversion of methane compared to the packed bed reactor, however the heat loss due to the membrane and the hydrogen leaving through the tube is decreasing the performances of the MR over the PBR compared to the isothermal case. Results show also that most of the reaction occurs at the very beginning of the reactor.

"The VIX was introduced in 1993 by the CBOE and has been commonly referred to as the fear gauge due to decreases in market sentiment leading market participants to purchase protection from declining asset prices. As market sentiment improves, declines in the VIX are generally observed. In reality the VIX measures the markets expectations about future volatility with asset prices either rising or falling in value. With the VIX gaining popularity in the marketplace a proliferation of derivative products has emerged allowing investors to trade volatility. In observance of the behavior of the VIX we attempt to model the derivative VXX as a mean reverting process via the Ornstein-Uhlenbeck stochastic differential equation. We extend this analysis by calibrating VIX options with observed market prices in order to extract the market density function. Using these parameters as the diffusion process in our Ornstein-Uhlenbeck model we derive futures prices on the VIX which serves to value our target derivative VXX."

"Characterizing the effects of subsurface wastewater effluent discharges remains as a significant challenge impacting both ground-water and surface water resources. Important aspects of this challenge relate to the quantification of the main processes affecting oxygen consumption within a wastewater plume and to the ability of representing these conditions over a range of scales. The goal of this research is to improve our understanding of the relevant processes affecting oxygen consumption and thus, controlling natural attenuation in wastewater contaminated aquifers, and also to characterize and quantify these processes through modeling approaches considering different scales. The analysis included consideration of restoration processes associated with a former sewage disposal discharge in Falmouth, MA. The discharge was removed in 1995, and the site has been experiencing natural restoration since removal. A small-scale natural gradient tracer test, completed 6 years after cessation of sewage disposal, was used in previous research to develop parameters to characterize aerobic respiration and nitrification processes, key oxygen consuming processes for this site. In addition, field monitoring by the United States Geological Survey has provided a series of concentration profiles at different locations along the flowpath associated with the contaminant source. For this research, predictions obtained with the existing model were used in conjunction with these concentration profiles to assess the sensitivity and applicability in the parameters from this small-scale test, as well as their pertinence to the larger scale restoration process. By evaluating the applicability of this model to different scales and the associated variability of key model parameters, the approach provided an improved characterization of the primary processes affecting oxygen consumption."

Ces travaux visaient à établir un protocole permettant de différencier des cellules pluripotentes (embryonnaires ou induites) en kératinocytes capables de générer un épiderme pluristratifié in vitro et in vivo pour la thérapie cellulaire et la modélisation pathologique. Dans une 1ère partie, des cellules hES ont été utilisées pour être différenciées en culture en kératinocytes. Ces cellules ont été caractérisées par des approches de RT-PCR quantitative, FACS et d’immunofluorescence. De plus, leur potentiel immunogène a été évalué par l’analyse en FACS des protéines du CMH de classe I et II. La capacité de ces kératinocytes à reformer un épiderme stratifié a été démontrée in vitro,sur une matrice synthétique, mais aussi in vivo après greffe sur souris immunodéficientes. Dans une 2ème partie, des cellules hIPS ont ensuite été dérivées en kératinocytes avec le même protocole. Si la caractérisation des kératinocytes issus des hIPS a produit des résultats positifs, la possibilité de maintenir ces cellules en culture s’est révélée impossible dans les conditions utilisées. De même, alors que la stratégie visait à modéliser les épidermolyses bulleuses, l’interférence avec l’expression de protéines impliquées dans la jonction épidermo-dermique n’a pas été couronnée de succès. Par contre, des nouvelles hIPS mutées ont pu être générées puis différenciées en kératinocytes, prouvant ainsi la possibilité d’utiliser cette démarche pour modéliser la pathologie. Dans une 3ème partie, une étude d’expression différentielle a permis de démontrer que les kératinocytes dérivés des cellules pluripotentes surexpriment la kératine 19 (marqueur foetal) par rapport à des kératinocytes postnataux. / This work aimed to establish a protocol for the differention of pluripotent cells (embryonic or induced) into keratinocytes, which must be able to generate a pluristratified epidermis in vitro as well as in vivo for cell therapy and pathological modeling. In the 1st part of this work, a new protocol was developped to derive keratinocytes from hES cells. These cells were characterized by using RTq-PCR, FACS and immunofluorescence assays. Their immunogenicity was also evaluated by FACS analysis of class I and class II MHC proteins. The capacity of these keratinocytes to generate a pluristratified epidermis has been proved in vitro as well as in vivo following grafts onto immunodeficient mice. In a 2nd part of this work, hIPS cells were then derived into keratinocytes using the same protocol. The characterization of heratinocytes derived from hIPS produced positive results, but it was unfortunately not possible to maintain those cells in culture. Similarly, a strategy based on RNA interference against dermoepidermal junction proteins was not successful; but new hIPS mutated cell lines have been established and differentiated into keratinocytes, proving the feasibility of such an approach for pathological modeling. In a 3rd part of this work, a differential expression study proved that keratinocytes derived from pluripotent cells overexpress keratin 19 (marker fetus) when compared with postnatal keratinocytes.

Bai Yun. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2006. / Includes bibliographical references (leaves 34-36). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- The Analysis of a Two-Level SEM with Group Specific Variables in LISREL --- p.4 / Chapter 2.1 --- The Model --- p.4 / Chapter 2.2 --- An Augmented Model --- p.5 / Chapter 2.3 --- Implementation in LISREL --- p.7 / Chapter 2.4 --- Simulation --- p.9 / Chapter 2.4.1 --- The Simulation Design --- p.9 / Chapter 2.4.2 --- Methods of Evaluation --- p.10 / Chapter 2.4.3 --- Simulation Results --- p.12 / Chapter 2.5 --- A Comparison to Mplus --- p.14 / Chapter 2.6 --- Empirical Demonstration: Multi-source Performance Appraisals --- p.14 / Chapter 3 --- Implementing Two level SEM with Cross-level Covariance Structures in Mx --- p.16 / Chapter 3.1 --- Two level Model Specifications with a Cross-level Covariance Structure --- p.17 / Chapter 3.2 --- An Illustrative Example --- p.20 / Chapter 3.3 --- Mx Simulation Design --- p.22 / Chapter 3.4 --- Simulation Results --- p.23 / Chapter 3.4.1 --- Accuracy of Parameter Estimates --- p.23 / Chapter 3.4.2 --- Accuracy of Standard Error Estimates --- p.24 / Chapter 3.4.3 --- Distribution of Goodness-of-fit Statistics --- p.24 / Chapter 3.5 --- Enlarged Mx Model --- p.24 / Chapter 3.5.1 --- Mx Model with Enlarged Xgi --- p.25 / Chapter 3.5.2 --- Mx Model with Enlarged Ng --- p.26 / Chapter 4 --- LISREL Sampling --- p.27 / Chapter 4.1 --- LISREL Sampling Simulation Design --- p.27 / Chapter 4.2 --- Simulation Results --- p.28 / Chapter 4.2.1 --- Accuracy of Parameter Estimates --- p.29 / Chapter 4.2.2 --- Accuracy of Standard Error Estimates --- p.30 / Chapter 4.2.3 --- Distribution of Goodness-of-fit Statistics --- p.30 / Chapter 5 --- Discussion --- p.31 / Appendices --- p.37 / Appendix 1 LISREL Sample Program --- p.37 / Appendix 2 LISREL Syntax for an ALL-Y Model --- p.38 / Appendix 3 LISREL Data Set Up --- p.39 / Appendix 4 Mx Sample Program --- p.40 / List of Figures / Chapter 1 --- The Augmented Two-level Model --- p.41 / Chapter 2 --- Results of the Performance Appraisal Example --- p.42 / Chapter 3 --- Two-level Model with a Cross-level Structure --- p.43 / Chapter 4 --- QQ-plot for P1-P6 --- p.44 / Chapter 5 --- QQ-plot for M1-M6 --- p.45 / List of Tables / Chapter 1 --- Simulation Conditions Associated with Each Pattern --- p.46 / Chapter 2 --- Simulation Results: Accuracy of Parameter Estimates --- p.47 / Chapter 3 --- Simulation Results: Precision of Standard Error Estimates --- p.48 / Chapter 4 --- Simulation Results: The Goodness-of-fit(GOF) Statistics --- p.49 / Chapter 5 --- Analysis of the Performance Appraisal Example --- p.49 / Chapter 6 --- Simulation Results: Mplus vs. LISREL-Parameter Estimates(l) --- p.50 / Chapter 7 --- Simulation Results: Mplus vs. LISREL-Parameter Estimates(2) --- p.51 / Chapter 8 --- Simulation Results: Mplus vs. LISREL-SE Estimates (Ratio) --- p.52 / Chapter 9 --- Simulation Results: Mplus vs. LISREL-GOF Statistics --- p.53 / Chapter 10 --- Mx Illustrative Example Results --- p.53 / Chapter 11 --- Mx Simulation Patterns --- p.53 / Chapter 12 --- Mx Simulation Results: Accuracy of Parameter Estimates --- p.54 / Chapter 13 --- Mx Simulation Results: MARB for Parameter and S.E. Estimates --- p.54 / Chapter 14 --- Mx Simulation Results: Goodness-of-fit Statistics --- p.55 / Chapter 15 --- Mx Simulation Results for M5 --- p.55 / Chapter 16 --- Mx Simulation Results for M5 and M6: Goodness-of-fit Statistics --- p.56 / Chapter 17 --- Mx Simulation Results for M6 --- p.56 / Chapter 18 --- LISREL Sampling: Simulation Patterns --- p.56 / Chapter 19 --- LISREL Sampling: Simulation Results for LI to L3 --- p.57 / Chapter 20 --- LISREL Sampling: Simulation Results for L4 to L6 --- p.58 / Chapter 21 --- LISREL Sampling: MARB for Parameter and S.E. Estimates --- p.59 / Chapter 22 --- LISREL Sampling: Goodness-of-fit Statistics --- p.59

This thesis focuses on the two main works that related in assessing the implications of (i) fiscal expansion (or contraction) and (ii) implementing a carbon tax on carbon-based fuels as well as the feed-in tariffs (subsidies to clean energy production) on Indonesia’s economy, within the context of static computable general equilibrium (CGE) analysis. In the first study, we investigate the impacts of increasing the public consumptions on Indonesia’s main macroeconomic indicators and to their consequences by examining how different institutions and sectors in the economy are affected. Three scenarios are carried out under different financing options to budgeting neutral the additional public spending. The results suggest that the increase of government expenditure on goods under the adjusted government saving generates the highest improvement on Indonesia’s GDP but results in a rise of budget deficit. In contrast, under the budget-neutral scheme of either reducing the subsidy rates across activities or increasing the output tax rates would result in less improvement to the Indonesia’s GDP. This is because a subsidy cut (or higher output tax) immediately escalates the production costs and, thus, increases the prices of final goods purchased by the households. These changes result in a fall of their real consumption that eventually leads to a drop in aggregate demand. However, compared to the scenario of subsidy cut, a higher output tax has the most adverse effects on national income. The industry’s production costs are more pressurized by a higher output tax. which in turn, creates deindustrialization, lower employment, and thus reduces the national income and output. In the second study, we investigate the two key frameworks to reduce Indonesia’s greenhouse gas (GHG) emissions: (i) implementing a carbon tax on fossil fuels; and (ii) promoting clean (renewable) energy production through the feed-in tariff (subsidy) scheme. In the carbon tax implementation, we assume that the government levies a tax of Rp. 100,000/ton CO2e with three possible revenue-recycling scenarios. In a first scenario, we allow the carbon tax to be recycled through adjustment of the labour (income) tax rate. In a second scenario, we allow the government to increase their spending on goods proportionally to compensate the revenue raised from a carbon tax. And finally, in the third scenario, we assume that the additional revenue from carbon tax is kept to run a budget surplus (government saving adjusts). Whilst, in the feed-in tariff (FIT) scenario, we assume that the government sets a 13.14% subsidy rate to renewable generations (hydro and geothermal generation) where the support payments are distributed equally among electricity consumers through a higher electricity tax rate. Overall, the results suggested that the carbon tax, in the short run, reduces the national emissions but raises costs to the economy, resulting a fall in GDP. In terms of income distribution, the carbon tax tends to be progressive in both (first and second) scenarios of revenue-recycling. However, when there is no compensating (recycling) mechanism (third scenario), the carbon tax tends to be regressive - the poorer households carry a higher share of the carbon tax burden. On the other hand, in case of the FIT scheme (15% subsidy to renewable generation), the impacts are negligible on national income and emissions. This is because the initial renewable shares in the electricity mix are small (a 11% share from hydro generation and a 5% share from geothermal generation); and these technology outputs are only utilized in the electricity industry. Therefore, we argue that the current Indonesia’s FIT regulation – about 13.14% subsidy rate for renewable generation technologies – is ineffective to reduce the national emissions.

La maîtrise de la QoS dans 802.11e EDCA (Enhanced Distributed Coordination Function) ne peut être assurée que par un mécanisme de contrôle d’admission qui empêche le réseau d’atteindre un état de saturation critique et par la même garantit les besoins de QoS des applications voix/vidéo. Ce mécanisme de contrôle d’admission a besoin pour sa prise de décision de prédire les métriques de performances si un nouveau flux est admis. Dans le but de rendre les décisions efficaces, nous choisissons d’utiliser une méthode de prédiction basée sur un modèle analytique. Ce dernier doit remplir deux conditions : 1) fournir une bonne précision de prédiction et 2) avoir une complexité numérique faible et un temps de réponse limité. Vu que la majorité des modèles analytiques de la littérature ne satisfont pas à ces deux conditions, nous développons un nouveau modèle analytique pour EDCA qui est capable de prédire le débit et le délai d’accès des différentes Access Category (AC) d’EDCA. Ainsi, après la modélisation analytique du temps de transmission des ACs en prenant en compte le paramètre de différentiation TXOPLimit, nous développons un modèle analytique pour EDCA sous la forme d’une chaîne de Markov à quatre dimensions. Celui-ci est développé d’abord dans les conditions de saturation puis étendu aux conditions générales de trafic. Pour finir, nous proposons un algorithme de contrôle d’admission à implémenter au sein du point d’accès et qui utilise le modèle analytique proposé. Nous proposons un abaque de solution d’optimisation des paramètres d’accès d’EDCA. Le but étant d’améliorer les performances du mécanisme de contrôle d’admission par l’utilisation optimale des ressources du réseau. / The QoS control in 802.11e EDCA (Enhanced Distributed Coordination Function) cannot be assured without an admission control mechanism which is capable of stopping the network from reaching a high saturation state and therefore guarantee the QoS requirements for voice and video applications. This admission control mechanism needs to predict the performance metrics that can be achieved by the network before deciding to admit any new flow. In order to obtain accurate decisions, we chose to use a prediction method based on an analytical model. The later must 1) grant the best accuracy of the prediction and 2) have a low computational complexity. Knowing that the current literatures’ major analytical models do not satisfy these two conditions, we therefore develop a new analytical model for EDCA capable to predict the achievable performance metrics of different Access Categories (ACs) of EDCA such as the throughput and access delay. Hence, after the analytical modeling of the transmission time of different ACs while taking into account the TXOPLimit differentiation parameter, we develop an analytical model for EDCA based on a four dimensional Markov Chain. This model is developed first in the saturation conditions and then extended to general traffic conditions. Finally, we propose the admission control algorithm to be implemented within the QoS Access Point (QAP) that uses the analytical model proposed. As a final point, we propose an abacus solution to optimize the configuration of EDCA access parameters. The objective is to enhance the performance of the admission control algorithm by the optimal use of network resources.

Master of Science / In this thesis, a recently developed physiologically-based model of the sleep-wake switch is analyzed and applied to a variety of clinically-relevant protocols. In contrast to phenomenological models, which have dominated sleep modeling in the past, the present work demonstrates the advantages of the physiologically-based approach. Dynamical and linear stability analyses of the Phillips-Robinson sleep model allow us to create a general framework for determining its response to arbitrary external stimuli. The effects of near-stable wake and sleep ghosts on the model’s dynamics are found to have implications for arousal during sleep, sleep deprivation, and sleep inertia. Impulsive sensory stimuli during sleep are modeled modeled according to their known physiological mechanism. The predicted arousal threshold variation matches experimental data from the literature. In simulating a sleep fragmentation protocol, the model simultaneously reproduces the body temperature and arousal threshold variation measured in another existing clinical study. In the second part of the thesis, we simulate sleep deprivation by introducing a wake-effort drive that is required to maintain wakefulness during normal sleeping periods. We interpret this drive both physiologically and psychologically, and demonstrate quantitative agreement between the model’s output and experimental subjective fatigue-related data. As well as subjective fatigue, the model is simultaneously able to reproduce adrenaline excretion and body temperature variations. In the final part of the thesis, the model is extended to include the orexinergic neurons of the lateral hypothalamic area. Due to the dynamics of the orexin group, the extended model exhibits sleep inertia, and an inhibitory circadian projection to the orexin group produces a postlunch dip in performance – both of which are well-known behavioral features. Including both homeostatic and circadian inputs to the orexin group, the model produces a waking arousal variation that quantitatively matches published clinical data.