Correct Specification Research Articles

Validation of GNSS baseline observation models using information criteria

This paper adapts and discusses the information criterion (IC) approach for the validation of baseline GNSS observation models in the context of outliers. Three IC methods are compared with the conventional procedure of iterative data snooping (IDS). To verify the reliability of the two considered approaches, several scenarios differing in: the magnitude and number of outliers, and the interval between measurement epochs were investigated. Generally, the IDS procedure achieved the highest rates of correct model specification when no outlier existed. However, as the number of outliers was increasing one of the IC methods was starting to achieve the highest rates.

Investigating the Applicability of Logistic Regression and Artificial Neural Networks in Predicting Breast Cancer

Breast cancer has become recently the most common cancer and a major cause of death among women all over the world and especially in developing countries like Iraq. This study aims to predict the type of breast tumor whether benign or malignant through the different models that were built using logistic regression and neural networks which is expected to be helpful for oncologists in diagnosing the type of breast tumor. Four models were set using binary logistic regression and two different types of artificial neural networks namely multilayer perceptron (MLP) and radial basis function (RBF). Both validated and trained models were evaluated using different performance metrics like accuracy or correct classification rate (CCR), receiver operating characteristic (ROC) curves, area under ROC curve (AUC), sensitivity and specificity. Dataset has been downloaded from the machine learning repository of University of California, Irvin (UCI ml repository) that consists of 9 attributes and 699 valid instances. 
 Firstly, some preprocessing was done to cleanse the data, then the models were built using the Logistic Regression method and Artificial Neural Networks and a comparison was done to find out which model will give the highest performance. Each model was validated with a different dataset than that used for developing the models. The analysis of the results showed that the Radial Basis Function neural network model is the best classifier in the prediction of the type of breast tumors since it had recorded the highest performance in terms of correct classification rate (accuracy), sensitivity, specificity, and AUC among all other models.

Two-step combined nonparametric likelihood estimation of misspecified semiparametric models

This paper proposes to estimate possibly misspecified semiparametric estimating equations models using a two-step combined nonparametric likelihood method. The method uses in the first step the plug in principle and replaces the infinite dimensional parameter with a consistent estimator. In the second step an estimator for the finite dimensional parameter is obtained by combining exponential tilting with a another member of the empirical Cressie-Read discrepancy. The resulting class of semiparametric estimators are robust to misspecification and have the same asymptotic variance as that of the efficient semiparametric generalised method of moment estimator under correct specification. It is also shown that the asymptotic distributions of the proposed estimators can be consistently estimated by a multiplier bootstrap procedure. The results of the paper are illustrated with a quadratic inference function model and an instrumental variable partially linear additive model. Monte Carlo evidence suggests that the proposed estimators have competitive finite sample properties.

The evaluation and comparison of three benchmark asset pricing models with daily data: supplementary evidence

ABSTRACT Recent studies advocate two new benchmark models (the Fama-French five-factor model and the Hou, Xue and Zhang four-factor model) with monthly data. Our daily data approach provides considerable supplement to the monthly data approach presented in recent studies. We adopt the advanced bootstrap methodology by replicating the original data sample, and this approach should effectively alleviate the problem of too much noise in the data of daily return. A two-pass cross-sectional regression and GMM with several useful testing statistics are used to more thoroughly diagnose the specifications of the model. The following consistency is observed when using different frequencies of sample data: the evidence indicates that the two newer benchmark models (the Fama-French five-factor model and the Hou, Xue and Zhang four-factor model) outperform the Fama-French three-factor model in estimating a few well-known portfolios (formed on different anomalies). However, several specification tests do not robustly accept the correct specifications of the Fama-French five-factor model and the Hou, Xue and Zhang four-factor model.

Causal inference and counterfactual prediction in machine learning for actionable healthcare

Big data, high-performance computing, and (deep) machine learning are increasingly becoming key to precision medicine—from identifying disease risks and taking preventive measures, to making diagnoses and personalizing treatment for individuals. Precision medicine, however, is not only about predicting risks and outcomes, but also about weighing interventions. Interventional clinical predictive models require the correct specification of cause and effect, and the calculation of so-called counterfactuals, that is, alternative scenarios. In biomedical research, observational studies are commonly affected by confounding and selection bias. Without robust assumptions, often requiring a priori domain knowledge, causal inference is not feasible. Data-driven prediction models are often mistakenly used to draw causal effects, but neither their parameters nor their predictions necessarily have a causal interpretation. Therefore, the premise that data-driven prediction models lead to trustable decisions/interventions for precision medicine is questionable. When pursuing intervention modelling, the bio-health informatics community needs to employ causal approaches and learn causal structures. Here we discuss how target trials (algorithmic emulation of randomized studies), transportability (the licence to transfer causal effects from one population to another) and prediction invariance (where a true causal model is contained in the set of all prediction models whose accuracy does not vary across different settings) are linchpins to developing and testing intervention models. Machine learning models are commonly used to predict risks and outcomes in biomedical research. But healthcare often requires information about cause–effect relations and alternative scenarios, that is, counterfactuals. Prosperi et al. discuss the importance of interventional and counterfactual models, as opposed to purely predictive models, in the context of precision medicine.

Acquisition of the Midbrain Dopaminergic Neuronal Identity.

The mesodiencephalic dopaminergic (mdDA) group of neurons comprises molecularly distinct subgroups, of which the substantia nigra (SN) and ventral tegmental area (VTA) are the best known, due to the selective degeneration of the SN during Parkinson’s disease. However, although significant research has been conducted on the molecular build-up of these subsets, much is still unknown about how these subsets develop and which factors are involved in this process. In this review, we aim to describe the life of an mdDA neuron, from specification in the floor plate to differentiation into the different subsets. All mdDA neurons are born in the mesodiencephalic floor plate under the influence of both SHH-signaling, important for floor plate patterning, and WNT-signaling, involved in establishing the progenitor pool and the start of the specification of mdDA neurons. Furthermore, transcription factors, like Ngn2, Ascl1, Lmx1a, and En1, and epigenetic factors, like Ezh2, are important in the correct specification of dopamine (DA) progenitors. Later during development, mdDA neurons are further subdivided into different molecular subsets by, amongst others, Otx2, involved in the specification of subsets in the VTA, and En1, Pitx3, Lmx1a, and WNT-signaling, involved in the specification of subsets in the SN. Interestingly, factors involved in early specification in the floor plate can serve a dual function and can also be involved in subset specification. Besides the mdDA group of neurons, other systems in the embryo contain different subsets, like the immune system. Interestingly, many factors involved in the development of mdDA neurons are similarly involved in immune system development and vice versa. This indicates that similar mechanisms are used in the development of these systems, and that knowledge about the development of the immune system may hold clues for the factors involved in the development of mdDA neurons, which may be used in culture protocols for cell replacement therapies.

Smooth coefficient estimation of stochastic frontier models

This paper proposes two alternative estimators for the semiparametric smooth coefficient stochastic frontier model which do not require parametric specification of the parameters of the distribution of inefficiency to identify all of the model primitives. These new estimators offer avenues for testing for correct specification. A small Monte Carlo simulation study reveals that the new methods perform similarly when correct specification is present and that the existing smooth coefficient estimator can perform poorly when it is incorrectly specified.

Secondary analysis of case-control association studies: Insights on weighting-based inference motivate a new specification test.

Case-control sampling is frequently used in genetic association studies to examine the relationship between disease and genetic exposures. Such designs usually collect extensive information on phenotypes beyond the primary disease, whose associations with the genetic exposures are also of great interest. Because the cases are over-sampled, appropriate analysis of secondary phenotypes should take into account this biased sampling design. We previously introduced a weighting-based estimator for appropriate secondary analysis, but have not thoroughly explored its statistical properties. In this article, we revisit our previous estimator to offer new insights and methodological extensions. Specifically, we extend our previous estimator and construct its more general form based on generalized least squares (GLS). Such an extension allows us to connect the GLS estimator with the generalized method of moments and motivates a new specification test designed to assess the adequacy of the disease model or the weights. The specification test statistic measures the weighted discrepancy between the case and control subsample estimators, and asymptotically follows a central Chi-squared distribution under correct disease model specification. We illustrate the GLS estimator and specification test using a case-control sample of peripheral arterial disease, and use simulations to further shed light on the operating characteristics of the specification test.

A replication of “A quasi-maximum likelihood approach for large, approximate dynamic factor models” (Review of Economics and Statistics, 2012)

Abstract The authors replicate and extend the Monte Carlo experiment presented in Doz, Giannone and Reichlin (A Quasi-Maximum Likelihood Approach For Large, Approximate Dynamic Factor Models, Review of Economics and Statistics, 2012) on alternative (time-domain based) methods for extracting dynamic factors from large datasets; they employ open source software and consider a larger number of replications and a wider set of scenarios. Their narrow sense replication exercise fully confirms the results in the original article. As for their extended replication experiment, the authors examine the relative performance of competing estimators under a wider array of cases, including richer dynamics, and find that maximum likelihood (ML) is often the dominant method; moreover, the persistence characteristics of the observable series play a crucial role and correct specification of the underlying dynamics is of paramount importance.

Robust estimation for longitudinal data under outcome‐dependent visit processes

SummaryIn longitudinal data where the timing and frequency of the measurement of outcomes may be associated with the value of the outcome, significant bias can occur. Previous results depended on correct specification of the outcome process and a somewhat unrealistic visit process model. In practice, this will never exactly be the case, so it is important to understand to what degree the results hold when those assumptions are violated in order to guide practical use of the methods. This paper presents theory and the results of simulation studies to extend our previous work to more realistic visit process models, as well as Poisson outcomes. We also assess the effects of several types of model misspecification. The estimated bias in these new settings generally mirrors the theoretical and simulation results of our previous work and provides confidence in using maximum likelihood methods in practice. Even when the assumptions about the outcome process did not hold, mixed effects models fit by maximum likelihood produced at most small bias in estimated regression coefficients, illustrating the robustness of these methods. This contrasts with generalised estimating equations approaches where bias increased in the settings of this paper. The analysis of data from a study of change in neurological outcomes following microsurgery for a brain arteriovenous malformation further illustrate the results.

Testing Measurement Invariance Across Unobserved Groups: The Role of Covariates in Factor Mixture Modeling.

Factor mixture modeling (FMM) has been increasingly used to investigate unobserved population heterogeneity. This study examined the issue of covariate effects with FMM in the context of measurement invariance testing. Specifically, the impact of excluding and misspecifying covariate effects on measurement invariance testing and class enumeration was investigated via Monte Carlo simulations. Data were generated based on FMM models with (1) a zero covariate effect, (2) a covariate effect on the latent class variable, and (3) covariate effects on both the latent class variable and the factor. For each population model, different analysis models that excluded or misspecified covariate effects were fitted. Results highlighted the importance of including proper covariates in measurement invariance testing and evidenced the utility of a model comparison approach in searching for the correct specification of covariate effects and the level of measurement invariance. This approach was demonstrated using an empirical data set. Implications for methodological and applied research are discussed.

Robust and efficient semi-supervised estimation of average treatment effects with application to electronic health records data.

We consider the problem of estimating the average treatment effect (ATE) in a semi-supervised learning setting, where a very small proportion of the entire set of observations are labeled with the true outcome but features predictive of the outcome are available among all observations. This problem arises, for example, when estimating treatment effects in electronic health records (EHR) data because gold-standard outcomes are often not directly observable from the records but are observed for a limited number of patients through small-scale manual chart review. We develop an imputation-based approach for estimating the ATE that is robust to misspecification of the imputation model. This effectively allows information from the predictive features to be safely leveraged to improve efficiency in estimating the ATE. The estimator is additionally doubly-robust in that it is consistent under correct specification of either an initial propensity score model or a baseline outcome model. It is also locally semiparametric efficient under an ideal semi-supervised model where the distribution of the unlabeled data is known. Simulations exhibit the efficiency and robustness of the proposed method compared to existing approaches in finite samples. We illustrate the method by comparing rates of treatment response to two biologic agents for treatment inflammatory bowel disease using EHR data from Partners'Healthcare.

Temporal progression of Drosophila medulla neuroblasts generates the transcription factor combination to control T1 neuron morphogenesis

Proper neural function depends on the correct specification of individual neural fates, controlled by combinations of neuronal transcription factors. Different neural types are sequentially generated by neural progenitors in a defined order, and this temporal patterning process can be controlled by Temporal Transcription Factors (TTFs) that form temporal cascades in neural progenitors. The Drosophila medulla, part of the visual processing center of the brain, contains more than 70 neural types generated by medulla neuroblasts which sequentially express several TTFs, including Homothorax (Hth), eyeless (Ey), Sloppy paired 1 and 2 (Slp), Dichaete (D) and Tailless (Tll). However, it is not clear how such a small number of TTFs could give rise to diverse combinations of neuronal transcription factors that specify a large number of medulla neuron types. Here we report how temporal patterning specifies one neural type, the T1 neuron. We show that the T1 neuron is the only medulla neuron type that expresses the combination of three transcription factors Ocelliless (Oc or Otd), Sox102F and Ets65A. Using CRISPR-Cas9 system, we show that each transcription factor is required for the correct morphogenesis of T1 neurons. Interestingly, Oc, Sox102F and Ets65A initiate expression in neurons beginning at different temporal stages and last in a few subsequent temporal stages. Oc expressing neurons are generated in the Ey, Slp and D stages; Sox102F expressing neurons are produced in the Slp and D stages; while Ets65A is expressed in subsets of medulla neurons born in the D and later stages. The TTF Ey, Slp or D is required to initiate the expression of Oc, Sox102F or Ets65A in neurons, respectively. Thus, the neurons expressing all three transcription factors are born in the D stage and become T1 neurons. In neurons where the three transcription factors do not overlap, each of the three transcription factors can act in combination with other neuronal transcription factors to specify different neural fates. We show that this way of expression regulation of neuronal transcription factors by temporal patterning can generate more possible combinations of transcription factors in neural progeny to diversify neural fates.

Multi‐kernel‐based random vector functional link network with decomposed features for epileptic EEG signal classification

This study proposes an improved hybrid model built with empirical mode decomposition (EMD) features combined with weighted multi-kernel random vector functional link network (WMKRVFLN) where the kernel parameters are optimised with an efficient optimisation algorithm known as water cycle algorithm (WCA) for diagnosis and classification of epileptic electroencephalogram (EEG) signals. The proposed model with optimisation is known as WCA–EMD–WMKRVFLN. The tanh and wavelet kernel functions are contributing together to the effectiveness of the proposed model. The features generated from EMD in terms of intrinsic mode functions (IMFs) are modulated to find important statistical and entropy based features and these features in a reduced form are employed as inputs to the model to classify epileptic EEG signals. The presented approach is evaluated in terms of percentage correct classification accuracy (ACC), specificity and sensitivity using two datasets and is compared with different classifiers and state-of-the-art techniques. The highest accuracies of 99.69% (five-class) and 100% (three-class) achieved using the Bonn-University dataset and 99.0% ACC (two-class) achieved using the Bern-Barcelona dataset. The achieved results report that the presented approach is a promising approach for EEG signal classification and is superior to several state-of-the-art techniques and is highly comparable to many such techniques.

Using Both Time Tradeoff and Discrete Choice Experiments in Valuing the EQ-5D: Impact of Model Misspecification on Value Sets.

Introduction. The EQ-5D-5L valuation protocol contains both time tradeoff (TTO) tasks and discrete choice experiments (DCE), raising the question of how to best use these in creating a value set. The hybrid model, which combines TTO and DCE data, has emerged as a commonly used approach. However, this model assumes independence among responses from the same individual, a linear relationship between TTO and DCE utilities, and, in many implementations, homoscedastic residuals. The aims of this study are to examine alternatives to these assumptions and determine the impact of misspecification on value sets. Methods. We performed a simulation study, parameterized using the US EQ-5D-5L valuation study, to assess the impact of model misspecification. We simulated TTO and DCE data with nonlinear relationships between TTO and DCE utilities, heteroscedastic errors, and correlated responses. Simulated data were analyzed using hybrid models with and without heteroscedasticity, Tobit models with and without heteroscedasticity, a latent class model, and a mixed model. Results. Mean absolute errors (MAEs) for correctly specified models were <0.05, whereas models that incorrectly assumed a linear relationship between TTO and DCE utilities or homoscedasticity of TTO responses featured states with an MAE >0.1. When a linear relationship between TTO and DCE utilities held, using both TTO and DCE data under correct specification yielded smaller MAEs compared with using TTO data alone but yielded larger MAEs when a linear relationship did not hold. Mistakenly assuming homoscedasticity led to increased MAEs, whereas ignoring dependence did not. Conclusions. Because heteroscedasticity in TTO utilities and nonlinear associations between DCE and TTO utilities have been noted, we recommend careful assessment of scedasticity and linearity to ascertain the suitability of a hybrid model.

Effect of Some Construction Factors of Bi-layer Knitted Fabrics Produced for Sports Wear on Resisting Ultraviolet Radiation

Ultraviolet rays represent a very low fraction of the solar spectrum which damages all living organisms and their metabolisms. Such radiations can cause a variety of symptoms, from basic tanning to extremely malignant skin cancers, fabrics with correct design specification, absorbers, and finishing methods, can play a very important role as a shielding system against the deterious effects of Ultraviolet rays. This study will stress on bi-layer knitted fabrics because knitted goods of all kinds are widely popular due to their flexibility ability to adjust, stretch when worn to a specific shape and their general comfortable wear. Knitted fabrics are widely used in Sportswear, summer clothes, under wears and Childs' clothes due to their excellent stretch and recovery, porosity, air permeability. Also knitted fabrics especially weft knitted fabrics are easier to be produced, and less cost compared to woven fabrics. Because of this fact, most previous research concerns woven structures that are typically less porous and offer higher protection against UV. The possibility of obtaining knitted fabrics with sufficient UV protection factor is of great interest, however, as knitted fabrics are more suitable for both sports and casual summer fashion garments. Current literature on the UV protection factor of knitted fabrics is very small and mostly concerns fabrics developed in machines with relatively large gauges This study is trying to find a solution for this problem by studying the Influence of weft knitted fabrics Construction with its' different parameters for Protection against Ultraviolet Radiation in sportswear applications.

Specification Tests for Generalized Propensity Scores Using Double Projections

This paper proposes a new class of nonparametric tests for the correct specification of generalized propensity score models. The test procedure is based on two different projection arguments, which lead to test statistics with several appealing properties. They accommodate high-dimensional covariates; are asymptotically invariant to the estimation method used to estimate the nuisance parameters and do not requite estimators to be root-n asymptotically linear; are fully data-driven and do not require tuning parameters, can be written in closed-form, facilitating the implementation of an easy-to-use multiplier bootstrap procedure. We show that our proposed tests are able to detect a broad class of local alternatives converging to the null at the parametric rate. Monte Carlo simulation studies indicate that our double projected tests have much higher power than other tests available in the literature, highlighting their practical appeal.

A verified and optimized Stream X‐Machine testing method, with application to cloud service certification

SummaryThe Stream X‐Machine (SXM) testing method provides strong and repeatable guarantees of functional correctness, up to a specification. These qualities make the method attractive for software certification, especially in the domain of brokered cloud services, where arbitrage seeks to substitute functionally equivalent services from alternative providers. However, practical obstacles include the difficulty in providing a correct specification, the translation of abstract paths into feasible concrete tests and the large size of generated test suites. We describe a novel SXM verification and testing method, which automatically checks specifications for completeness and determinism, prior to generating complete test suites with full grounding information. Three optimization steps achieve up to a 10‐fold reduction in the size of the test suite, removing infeasible and redundant tests. The method is backed by a set of tools to validate and verify the SXM specification, generate technology‐agnostic test suites and ground these in SOAP, REST or rich‐client service implementations. The method was initially validated using seven specifications, three cloud platforms and five grounding strategies.

Reasonableness as a virtue of citizenship and the opacity respect requirement

This article defends a specific account of reasonableness as a virtue of liberal citizenship. I specify an account of reasonableness that I argue is more consistent with the phenomenology of intersubjective exchanges among citizens over political matters in contexts of deep disagreement. My reading requires reasonable citizens to undertake an attitude of epistemic modesty while deliberating public matters with agents who hold views different from theirs. In contrast with my view, I debate Martha Nussbaum’s and Steven Wall’s accounts of reasonableness and specify why I believe that these proposals, although interesting, both require revisions. Distinguishing my account from theirs, I specify the normative relation between reasonableness and a general framework of political legitimacy that identifies citizens as ‘co-authors of democratic decisions’. Here, I argue that the liberal ideal of ascribing to each member of the constituency the status of putative epistemic authority can be properly fulfilled if coupled with a correct specification of the political ideal of mutual respect. I conclude claiming that opacity respect, a notion of respect according to which the recognition respect that is owed to individuals is expressed by the idea that we have to treat them as ‘opaque’, is the most adequate concept of political respect when dealing with interpersonal deliberations at political level in contexts of deep disagreement.

Sampling Procedure for Establishing Criteria for Sea Urchin (Loxechinus albus) Proportion under the Legal Minimum Length of landings

Sea urchin (Loxechinus albus) is one of the most economically important species in the northeast of Chilean Patagonia, forming part of the highly diverse benthic community. This resource is being harvested under selective fishing pressure, which suppresses growth rates. In response, the National Standards Institute established Regulation 44 as a quality control measure, based on a sampling approach considering the percentage of sea urchin under the minimum legal size (UMLS) through an acceptance sampling plan. In 2019, the Chilean Undersecretary of Fisheries and Aquaculture (SUBPESCA), under Exempt Resolution No. 910, ruled that any vessel that landed 15% with sea urchin of 6.5 cm UMLS of the total catch weight will be sanctioned. In this paper, a sampling design is presented to establish criteria regarding compliance with this provision. The plan provides a statistical methodology for the evaluation of the percentage of UMLS sea urchin, which corroborates and certifies the current legislation that sanctions and fines the capture of sea urchin that does not comply with UMLS values outlined in Regulation 44. Assuming a systematic sampling plan and under assumed UMLS values of 6.0, 6.5, and 7.0 cm, it is determined that 15% of total landings (margin of tolerance) of sea urchin under 6.5 cm is the optimal level. Thus, a correct specification of the UMLS allows protecting the most vulnerable fraction of the population from overexploitation for growth.