Structural Model Parameters Research Articles

Assessing the treatment effect in a randomized controlled trial with extensive non-adherence: the EVOLVE trial.

Intention-to-treat (ITT) analysis is widely used to establish efficacy in randomized clinical trials. However, in a long-term outcomes study where non-adherence to study drug is substantial, the on-treatment effect of the study drug may be underestimated using the ITT analysis. The analyses presented herein are from the EVOLVE trial, a double-blind, placebo-controlled, event-driven cardiovascular outcomes study conducted to assess whether a treatment regimen including cinacalcet compared with placebo in addition to other conventional therapies reduces the risk of mortality and major cardiovascular events in patients receiving hemodialysis with secondary hyperparathyroidism. Pre-specified sensitivity analyses were performed to assess the impact of non-adherence on the estimated effect of cinacalcet. These analyses included lag-censoring, inverse probability of censoring weights (IPCW), rank preserving structural failure time model (RPSFTM) and iterative parameter estimation (IPE). The relative hazard (cinacalcet versus placebo) of mortality and major cardiovascular events was 0.93 (95% confidence interval 0.85, 1.02) using the ITT analysis; 0.85 (0.76, 0.95) using lag-censoring analysis; 0.81 (0.70, 0.92) using IPCW; 0.85 (0.66, 1.04) using RPSFTM and 0.85 (0.75, 0.96) using IPE. These analyses, while not providing definitive evidence, suggest that the intervention may have an effect while subjects are receiving treatment. The ITT method remains the established method to evaluate efficacy of a new treatment; however, additional analyses should be considered to assess the on-treatment effect when substantial non-adherence to study drug is expected or observed.

Efficient In-Plane Tire Mode Identification by Radial-Tangential Eigenvector Compounding

ABSTRACT Tire modal testing is frequently used for validation of numerical tire models and identification of structural tire model parameters. Most studies focus primarily on in-plane dynamic tire behavior and adopt the approach of the fixed boundary condition at the wheel center. Here, an identification method of in-plane tire dynamics was developed for the case of a free tire-rim combination. This particular case is important when the aim is to construct a full tire model, capable of predicting ride and noise, vibration, and harshness involving the whole vehicle, all from modal testing. Key attributes of the proposed approach include ease of implementation and efficient processing of measurements. For each type of excitation, i.e., radial and tangential, both radial and tangential responses were recorded. Compounding of the corresponding radial/tangential eigenvectors, which, in the context of the present work, refers to expressing the motion of the tire belt as a combination of the radial and tangential responses, results in smooth mode shapes that were found to agree with those published in other analytical and experimental studies.

Derivation of a new parametric impulse response matrix utilized for nodal wind load identification by response measurement

This paper provides new formulations to derive the impulse response matrix, which is then used in the problem of load identification with application to wind induced vibration. The applied loads are inversely identified based on the measured structural responses by solving the associated discrete ill-posed problem. To this end — based on an existing parametric structural model — the impulse response functions of acceleration, velocity and displacement have been computed. Time discretization of convolution integral has been implemented according to an existing and a newly proposed procedure, which differ in the numerical integration methods. The former was evaluated based on a constant rectangular approximation of the sampled data and impulse response function in a number of steps corresponding to the sampling rate, while the latter interpolates the sampled data in an arbitrary number of sub-steps and then integrates over the sub-steps and steps. The identification procedure was implemented for a simulation example as well as an experimental laboratory case. The ill-conditioning of the impulse response matrix made it necessary to use Tikhonov regularization to recover the applied force from noise polluted measured response. The optimal regularization parameter has been obtained by L-curve and GCV method. The results of simulation represent good agreement between identified and measured force. In the experiments the identification results based on the measured displacement as well as acceleration are provided. Further it is shown that the accuracy of experimentally identified load depends on the sensitivity of measurement instruments over the different frequency ranges.

Constitutive model fingerprints in medium-amplitude oscillatory shear

Rheologists have expectations for signatures of linear viscoelastic properties, such as shapes of G′(ω) and G″(ω). Medium amplitude (or asymptotically-nonlinear) oscillatory shear (MAOS) provides additional nonlinear rheological information with low dimensional, well-defined material functions [Ewoldt and Bharadwaj, Rheol. Acta 52, 201–209 (2013)]. Here, we develop expectations of signatures (or fingerprints) for the four asymptotically-nonlinear material functions associated with MAOS, [e1](ω), [e3](ω), [v1](ω), [v3](ω). Although the linear fingerprints may be identical for different models, the asymptotically-nonlinear fingerprints may be different in magnitude, frequency-scaling, curve shapes, and sign changes. To perform the analysis, we collect/translate a library of available analytical strain-controlled MAOS fingerprints for seven different constitutive models. Using this library, we identify general trends and highlight key differences of asymptotic-nonlinear viscoelasticity. Asymptotic nonlinearities for all models considered here obey the terminal regime inter-relations and frequency scaling predicted by Bharadwaj and Ewoldt [J. Rheol. 58, 891–910 (2014)]. Unlike the positive linear viscoelastic measures, at least one of the four asymptotic nonlinearities changes signs with Deborah number (De). Following sign interpretations of Ewoldt and Bharadwaj [Rheol. Acta 52, 201–209 (2013)], we show that nonlinearities tend to be driven by strain-rates at small De, and by strains at large De, a trend observed for nearly all the constitutive models studied here, the exception being the model for dilute rigid dumbbell suspensions of Bird et al. [J. Chem. Phys. 140, 074904 (2014)]. Some constitutive models exhibit multiple sign changes at intermediate De and there may be no universal behavior of asymptotically-nonlinear fingerprints in this regime. Therefore, frequency-dependent signatures can be material-specific. This will allow inverse problems to infer structure, select models, and fit model parameters using asymptotically-nonlinear signatures. To illustrate this aspect, we demonstrate a fingerprint matching exercise with experimental measurements on a transiently cross-linked hydrogel system. We find that currently available model fingerprints can match the qualitative magnitudes and frequency dependence, but not the signs of the experimental transient network response.

On Bayesian estimation of marginal structural models.

The purpose of inverse probability of treatment (IPT) weighting in estimation of marginal treatment effects is to construct a pseudo-population without imbalances in measured covariates, thus removing the effects of confounding and informative censoring when performing inference. In this article, we formalize the notion of such a pseudo-population as a data generating mechanism with particular characteristics, and show that this leads to a natural Bayesian interpretation of IPT weighted estimation. Using this interpretation, we are able to propose the first fully Bayesian procedure for estimating parameters of marginal structural models using an IPT weighting. Our approach suggests that the weights should be derived from the posterior predictive treatment assignment and censoring probabilities, answering the question of whether and how the uncertainty in the estimation of the weights should be incorporated in Bayesian inference of marginal treatment effects. The proposed approach is compared to existing methods in simulated data, and applied to an analysis of the Canadian Co-infection Cohort.

Structural Design and Construction of Complex-Shaped Tall Buildings

Today's diverse architectural design trends have produced tall buildings of various forms, such as twisted, tilted, and free forms, as can be seen in the Infinity Tower in Dubai, Veer Towers in Las Vegas and Phare Tower project in Paris. This paper studies various structural system design options for these complex-shaped tall buildings and their performances. For each complex form category, tall buildings are designed with different contemporary structural systems, such as diagrids, braced tubes and outrigger systems. The structural efficiency of each system, in conjunction with building forms, heights and height-to-width aspect ratios, is studied. Parametric structural models are generated to investigate the impacts of variation of important geometric configurations of complex-shaped tall buildings. The models are then exported to structural engineering software for analyses and design. Based on the study results, comparative structural efficiency of different structural systems for tall buildings of each complex form category is estimated.

Matrix Powers in Spatial Econometric Models

This paper demonstrates several implications of the Cayley-Hamilton Theorem for specification, estimation, and identification of spatial autoregressive models. First, the minimal polynomial of the spatial weights matrix implies an upper bound on the admissible lag order in spatial autoregressive models when powers of the weights matrix are used to form higher-order spatial lags. Second, knowledge of the minimal polynomial facilitates the derivation of closed-form solutions of the marginal effects. Compared to the common practice of numerically inverting the spatial lag polynomial, the use of closed-form solutions enhances the interpretation of the structural model parameters and can be computationally attractive in large samples. Moreover, I demonstrate that the numerical inversion approach may result in biased finite-sample estimates of the marginal effects due to the sample size dependence of the spatial weights matrix even if all coefficient estimates are unbiased. The closed-form solutions can be used to obtain an exact expression of the bias and to avoid it in the first place. Third, I show that the matrix exponential spatial specification has a restricted spatial autoregressive model representation. Finally, I highlight that the order of the minimal polynomial of the spatial weights (or social interaction) matrix is related to the identification of endogenous and exogenous interaction effects in social networks.

Reducing complexity: An iterative strategy for parameter determination in biological networks

The dynamics of biological networks are fundamental to a variety of processes in many areas of biology and medicine. Understanding of such networks on a systemic level is facilitated by mathematical models describing these networks. However, since mathematical models of signalling networks commonly aim to describe several highly connected biological quantities and many model parameters cannot be measured directly, quantitative dynamic models often present challenges with respect to model calibration. Here, we propose an iterative fitting routine to decompose the problem of fitting a system of coupled ordinary differential equations describing a signalling network into smaller subproblems. Parameters for each differential equation are estimated separately using a Differential Evolution algorithm while all other dynamic quantities in the model are treated as input to the system. The performance of this algorithm is evaluated on artificial networks with known structure and known model parameters and compared to a conventional optimisation procedure for the same problem. Our analysis indicates that the procedure results in a significantly higher quality of fit and more efficient reconstruction of the true parameters than the conventional algorithm.

Controlling for Multilateral Resistance Terms in Linearized Trade Gravity Equations Without Spatial Econometrics

This paper studies linearized spatial structural gravity models of bilateral trade a la Behrens, Ertur and Koch (2012). We show that these models do in fact not require spatial econometric methods for estimation. This result follows from the nature of the specific spatial weights matrix, and from the exporter- or importerspecific nature of some regressors and the approximation error. All structural model parameters are identified from a linear regression that uses a spatial lag of the dependent variable as a control function.

A computational study on the interaction between O$_{2}$ and pristine and Ge-doped aluminum phosphide nanotubes

In this research the interaction between oxygen molecules and the outside and inside surfaces of pristine and Ge-doped (4,4) armchair and (6,0) zigzag models of aluminum phosphide nanotubes (AlPNTs) was systematically investigated using density function theory. Structural parameters, adsorption energy, quantum parameters, HOMO/LUMO orbitals, and nuclear quadrupole resonance (NQR) parameters were calculated for all models of AlPNTs. The aim of this work was to investigate the effects of Ge doping and O$_{2}$ adsorption on the electrical and structural parameters of (4,4) armchair and (8,0) zigzag models of AlPNTs. The results revealed that adsorption energies for all models were negative with exothermic chemical bonding. By doping Ge in spite of the B52 site of (4,4) armchair and (6,0) zigzag models the adsorption energy increased significantly from pristine values and therefore Ge doping increased the reactivity of the nanotubes to O$_{2}$ adsorption. The NQR results showed that in AlPNTs Al atoms at the edges of nanotubes played a significant role in determining the electronic behaviors of AlPNTs and the average values of C$_{Q}$ ($^{27}$Al) and $\eta_{Q}$ for the O$_{2}$ attached on (4,4) armchair and (6,0) zigzag AlPNTs were higher than those of the pristine model. Analysis of the electronic properties indicated that adsorption of O$_{2}$ reduced the energy gap of AlPNTs. Quantum molecular results showed that the global hardness ($\eta )$ of Ge-doped models was smaller than that of other models.

‘Structural Performance of Two-Hinged Islamic Arches’

‘Structural Performance of Two-Hinged Islamic Arches’ Shehdeh Ghannam, Husam Najm Abstract This paper presents an investigation of Islamic arch forms and structural performance based on its geometrical formation. Arch form depends on several interrelated variables such arc angles, the ratio of arc radius to arc span, and the height of the arch. These variables are controlled in Islamic architecture by the geometry. This paper provides insights for the geometric design of Islamic arches and outlines its structural analysis under different load cases. A parametric study of the buckling behavior of arches for several load cases, horizontal thrust, and bending moment will be analyzed. This paper aims to provide an understanding of the structural concepts of the arch form in Islamic architecture based on the significance of geometry in construction of Islamic arches. Arches described by geometry and proportions, plays a crucial role in the composition and structure of the arch form. Typically design parameters are based on estimating force vectors to determine arch shape, assuming that they represent the architecture properly and are constant. To describe arch architecture in more detail and compare geometrical and non-geometrical states, geometric approach was validated. Arch forms are constructed from historical sections, based on geometric language, as well as the shape of the arch. The proposed approach may help to identify functional structural concepts within the arch, with homogeneous architectural geometric properties. The related results are discussed regarding the parameters essential for structural arch modeling and challenge aesthetic parameters of the arch-model. Full Text: PDF DOI: 10.15640/jea.v3n1a5

A Dynamic Network Model of the Unsecured Interbank Lending Market

We introduce a structural dynamic network model of the formation of lending relationships in the unsecured interbank market. Banks are subject to random liquidity shocks and can form links with potential trading partners to bilaterally Nash bargain about loan conditions. To reduce credit risk uncertainty, banks can engage in costly peer monitoring of counterparties. We estimate the structural model parameters by indirect inference using network statistics of the Dutch interbank market from 2008 to 2011. The estimated model accurately explains the high sparsity and stability of the lending network. In particular, peer monitoring and credit risk uncertainty are key factors in the formation of stable interbank lending relationships that are associated with improved credit conditions. Moreover, the estimated degree distribution of the lending network is highly skewed with a few very interconnected core banks and many peripheral banks that trade mainly with core banks. Shocks to credit risk uncertainty can lead to extended periods of low market activity, amplified by a reduction in peer monitoring. Finally, our monetary policy analysis shows that a wider interest rate corridor leads to a more active market through a direct effect on the outside options and an indirect multiplier effect by increasing banks’ monitoring and search efforts.

Heterogeneity Analysis of Cucumber Canopy in the Solar Greenhouse

Detailed analysis of canopy structural heterogeneity is an essential step in conducting parameters for a canopy structural model. This paper aims to analyze the structural heterogeneity of a cucumber (Cucumis sativus L.) canopy by means of analyzing leaf distribution in a greenhouse environment with natural sunlight and also to assess the effect of structural canopy heterogeneity on light interception and photosynthesis. Two experiments and four measurements were carried out in autumn 2011 and spring 2012. A static virtual three-dimensional (3D) canopy structure was reconstructed using a 3D digitizing method. The diurnal variation of photosynthesis rate was measured using CIRAS-2 photosynthesis system. The results showed that, leaf azimuth as tested with the Rayleigh-test was homogeneous at vine tip over stage but turned heterogeneous at fruit harvest stage. After eliminating the influence of the environment on the azimuth using the von Mises-Fisher method, the angle between two successive leaves was 144°; at the same time, a rule for the azimuth distribution in the canopy was established, stating that the azimuth distribution in cucumber followed a law which was positive spin and anti-spin. Leaf elevation angle of south-oriented leaves was on average 13.8° higher than that of north-oriented leaves. The horizontal distribution of light interception and photosynthesis differed significantly between differently oriented leaves. East- and west-oriented leaves exhibited the highest photosynthetic rate. In conclusion, detailed analysis of canopy structural heterogeneity in this study indicated that leaf azimuth and elevation angle were heterogeneous in cucumber canopy and they should be explicitly described as they have a great impact both on light distribution and photosynthesis.

Immunization of Fixed-Income Portfolios Using an Exponential Parametric Model

Litterman e Scheinkman (1991) mostram que mesmo uma carteira de renda fixa duration imunizada (neutra) pode sofrer grandes perdas e, portanto, propõem fazer hedge de carteiras utilizando a análise de componentes principais. O problema é que tal abordagem só é possível quando as taxas de juros são observáveis. Assim, quando as taxas de juros não são observáveis, como é o caso da maior parte dos mercados de dívida externa e interna de diversos países emergentes, o método não é diretamente aplicável. Este artigo propõe uma abordagem alternativa que é realizar o hedge baseado nos fatores de um modelo paramétrico da estrutura a termo. A imunização seguindo esta abordagem não só se mostra simples como também equivalente ao procedimento proposto por Litterman e Scheinkman quando as taxas são observáveis. Exemplos do método para operações de hedge e alavancagem nos mercados de títulos de dívida pública brasileira indexada a inflação são apresentados. O trabalho ainda discute como montar a apreçar carteiras que replicam fatores de risco, o que permite extrair alguma informação sobre a expectativa dos agentes acerca do comportamento futuro da curva de juros.

Learning from Learners

Traditional macroeconomic learning algorithms are misspecified when all agents are learning simultaneously. In this paper, we produce a number of learning algorithms that do not share this failing, and show that this enables them to learn almost any solution, for any parameters, implying learning cannot be used for equilibrium selection. As a by-product, we are able to show that when all agents are learning by traditional methods, all deep structural parameters of standard new-Keynesian models are identified, overturning a key result of Cochrane (2009; 2011). This holds irrespective of whether the central bank is following the Taylor principle, irrespective of whether the implied path is or is not explosive, and irrespective of whether agents’ beliefs converge. If shocks are observed then this result is trivial, so following Cochrane (2009) our analysis is carried out in the more plausible case in which agents do not observe shocks.

COMPARATIVE EFFICIENCY BETWEEN STRUCTURAL SYSTEMS FOR COMPLEX-SHAPED TALL BUILDINGS

This paper presents various structural system design options for complex-shaped tall buildings—such as twisted, tilted, and freeform tall buildings—and evaluates their comparative structural efficiency. For each complex form category, tall buildings are designed with different contemporary tall building structural systems, e.g., diagrids, braced tubes, and outrigger systems. Comparative structural efficiency between these systems in conjunction with building forms and height is studied. The heights of the studied buildings range from 60 to 100 stories, and the corresponding height-to-width aspect ratios range from 6.5 to 10.8. Parametric structural models are generated with Rhino/Grasshopper to investigate the impacts of various important geometric configurations of complex-shaped tall buildings, such as the rate of twist, angle of tilt, and degree of fluctuation of free form. The parametric models are exported to structural engineering software SAP2000 for analyses and design. Based on the study results, comparative structural efficiency between different structural systems of each complex form category is presented.

MODELING MOVEMENTS IN INDIVIDUAL CONSUMPTION: A TIME-SERIES ANALYSIS OF GROUPED DATA

We characterize the time-series properties of group-level consumption, income, and interest rates using microdata. We relate the coefficients of moving average representations to structural parameters of theoretical models of consumption behavior. Using long time series of cross sections to construct synthetic panel data for the United Kingdom, we find that for high-educated individuals the Euler equation restrictions are not rejected, the elasticity of intertemporal substitution is higher than one, and there is evidence of “excess smoothness” of consumption. Low-educated individuals, conversely, exhibit excess sensitivity of consumption to past income, and the elasticity of intertemporal substitution is not statistically different from zero.

Modified Profile Likelihood for Fixed-Effects Panel Data Models

We show how modified profile likelihood methods, developed in the statistical literature, may be effectively applied to estimate the structural parameters of econometric models for panel data, with a remarkable reduction of bias with respect to ordinary likelihood methods. Initially, the implementation of these methods is illustrated for general models for panel data including individual-specific fixed effects and then, in more detail, for the truncated linear regression model and dynamic regression models for binary data formulated along with different specifications. Simulation studies show the good behavior of the inference based on the modified profile likelihood, even when compared to an ideal, although infeasible, procedure (in which the fixed effects are known) and also to alternative estimators existing in the econometric literature. The proposed estimation methods are implemented in an R package that we make available to the reader.

Ensemble learning of inverse probability weights for marginal structural modeling in large observational datasets.

Inverse probability weights used to fit marginal structural models are typically estimated using logistic regression. However, a data-adaptive procedure may be able to better exploit information available in measured covariates. By combining predictions from multiple algorithms, ensemble learning offers an alternative to logistic regression modeling to further reduce bias in estimated marginal structural model parameters. We describe the application of two ensemble learning approaches to estimating stabilized weights: super learning (SL), an ensemble machine learning approach that relies on V-fold cross validation, and an ensemble learner (EL) that creates a single partition of the data into training and validation sets. Longitudinal data from two multicenter cohort studies in Spain (CoRIS and CoRIS-MD) were analyzed to estimate the mortality hazard ratio for initiation versus no initiation of combined antiretroviral therapy among HIV positive subjects. Both ensemble approaches produced hazard ratio estimates further away from the null, and with tighter confidence intervals, than logistic regression modeling. Computation time for EL was less than half that of SL. We conclude that ensemble learning using a library of diverse candidate algorithms offers an alternative to parametric modeling of inverse probability weights when fitting marginal structural models. With large datasets, EL provides a rich search over the solution space in less time than SL with comparable results.

Abstract CT330: Adjusting for confounding effects of treatment crossover in a randomized phase 2 study of dabrafenib plus trametinib in BRAF V600+ metastatic melanoma

Abstract In a randomized phase 2 study (BRF113220), dabrafenib (150 mg twice daily) plus trametinib (2 mg once daily) improved progression-free survival (HR=0.39; 95% CI, 0.25-0.62; P<.001) vs dabrafenib alone. 54 patients were randomized to each arm. At the most recent data cut (March 2013), median duration of follow-up was 24 months and 52% of patients had died. ITT analyses of overall survival (OS) are likely to underestimate the true benefit of treatment because patients in the control arm (83%) were able to cross over to combination treatment upon progression. ITT median OS, a secondary endpoint, was 23.8 vs 20.2 months (HR=0.73; 95% CI, 0.43-1.24). We adjusted for confounding due to this crossover using randomization-based crossover adjustment methods: the rank-preserving structural failure time model (RPSFTM) and iterative parameter estimation (IPE) algorithm. Different assumptions about the durability of treatment effect were tested: “treatment group” analyses allow for a treatment effect that continues beyond treatment discontinuation, while “on treatment_observed” analyses assumed that the treatment effect disappears upon discontinuation. Results are presented in the Table. Reestimation of median OS following adjustment with both RPSFTM and IPE (possible only for the treatment group analyses) suggested an OS gain of 9.1 months with combination treatment vs dabrafenib alone (compared with a 3.6-month gain under ITT assumptions). The RPSFTM estimation method performed marginally better than the IPE analyses and may provide more robust results. After adjusting for the effect of crossover, the reduction in HRs to 0.47-0.50 suggests that the ITT analysis underestimates the effect of dabrafenib plus trametinib on OS. Results are uncertain due to the limiting assumptions associated with the RPSFTM and IPE methodology, small trial size, low number of observed deaths, and low number of control group patients who did not cross over. Crossover Adjustment MethodAssumptionOS HR (95% CI)None (ITT analysis unadjusted for crossover)0.73 (0.43-1.24)RPSFTMTreatment group0.47 (0.13-1.66)On treatment–observed analyses0.49 (0.15-1.63)IPETreatment group0.50 (0.16-1.54)On treatment–observed analyses0.47 (0.14-1.61) Citation Format: Nicholas Latimer, Mayur Amonkar, Ceilidh Stapelkamp, Peng Sun. Adjusting for confounding effects of treatment crossover in a randomized phase 2 study of dabrafenib plus trametinib in BRAF V600+ metastatic melanoma. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr CT330. doi:10.1158/1538-7445.AM2014-CT330