Statistical Inefficiency Research Articles

The slice sampler and centrally symmetric distributions

Abstract We show that the slice sampler generates Markov chains whose variables are mean independent and thus uncorrelated when the target density is centrally symmetric. Skewness instead boosts correlations. Popular implementation algorithms such as stepping-out and multivariate-sampling-with-hyperrectangles add statistical inefficiency, the first in case of multimodality, the second in all circumstances. A new sampler which exploits these structural and algorithmic characteristics to reduce the variance of Monte Carlo estimates is experimented in several sampling problems. An insight into the properties of the product slice sampler is also provided.

Enhanced Sampling in Molecular Dynamics Simulations: How Many MD Snapshots can be Needed to Reproduce the Biological Behavior?

Since its early days in the 19th century, medicinal chemistry has concentrated its efforts on the treatment of diseases, using tools from areas such as chemistry, pharmacology, and molecular biology. The understanding of biological mechanisms and signaling pathways is crucial information for the development of potential agents for the treatment of diseases mainly because they are such complex processes. Given the limitations that the experimental approach presents, computational chemistry is a valuable alternative for the study of these systems and their behavior. Thus, classical molecular dynamics, based on Newton's laws, is considered a technique of great accuracy, when appropriated force fields are used, and provides satisfactory contributions to the scientific community. However, as many configurations are generated in a large MD simulation, methods such as Statistical Inefficiency and Optimal Wavelet Signal Compression Algorithm are great tools that can reduce the number of subsequent QM calculations. Accordingly, this review aims to briefly discuss the importance and relevance of medicinal chemistry allied to computational chemistry as well as to present a case study where, through a molecular dynamics simulation of AMPK protein (50 ns) and explicit solvent (TIP3P model), a minimum number of snapshots necessary to describe the oscillation profile of the protein behavior was proposed. For this purpose, the RMSD calculation, together with the sophisticated OWSCA method was used to propose the minimum number of snapshots.

Acceleration of Solvation Free Energy Calculation via Thermodynamic Integration Coupled with Gaussian Process Regression and Improved Gelman-Rubin Convergence Diagnostics.

The determination of the solvation free energy of ions and molecules holds profound importance across a spectrum of applications spanning chemistry, biology, energy storage, and the environment. Molecular dynamics simulations are powerful tools for computing this critical parameter. Nevertheless, the accurate and efficient calculation of the solvation free energy becomes a formidable endeavor when dealing with complex systems characterized by potent Coulombic interactions and sluggish ion dynamics and, consequently, slow transition across various metastable states. In the present study, we expose limitations stemming from the conventional calculation of the statistical inefficiency g in the thermodynamic integration method, a factor that can hinder the determination of convergence of the solvation free energy and its associated uncertainty. Instead, we propose a robust scheme based on Gelman-Rubin convergence diagnostics. We leverage this improved estimation of uncertainties to introduce an innovative accelerated thermodynamic integration method based on the Gaussian Process regression. This methodology is applied to the calculation of the solvation free energy of trivalent rare-earth elements immersed in ionic liquids, a scenario in which the aforementioned challenges render standard approaches ineffective. The proposed method proves to be effective in computing solvation free energy in situations where traditional thermodynamic integration methods fall short.

A stochastic frontier analysis of cost of efficiency in Turkish cement industry

Purpose - The main focus of this paper is to measure cost efficiency in Turkish Cement Industry using firm level data over the period 2008-2016. The measurement of cost efficiency in the industry in which there can seem usually competition infringement is of great importance regarding effective allocation of resources. The ineffective allocation of resources can bring about cost inefficiency. Methodology - In this study, efficiency is measured using a frontier analysis approach, which quantifies efficiency as the gap between the best achievable performance and actual performance. The Stochastic Frontier Analysis (SFA) method is employed to assess cost efficiency of each firm in this manner. By incorporating both statistical error and inefficiency, the SFA method recognizes that deviations from optimal performance can arise from both random errors and inefficiency factors. Findings- Empirical results show that cost efficiency scores vary between 82% and 93% by different models. The results also indicate a wide range of cost efficiencies across different size cement firms. The lowest cost efficiency score is estimated by TFE, BC95, and PL81 models in 2015. Conclusion- Average cost efficiency score varies between approximately 93% and 82%, which means that the firms can potentially achieve at least 7% and a maximum of 18% reduction in input costs. Keywords: Stochastic frontier analysis, cement industry, cost efficiency, efficiency estimation, time-varying, invariant efficiency. JEL Codes: D24, C51, L61

Mitigating pseudoreplication and bias in resource selection functions with autocorrelation‐informed weighting

Abstract Resource selection functions (RSFs) are among the most commonly used statistical tools in both basic and applied animal ecology. They are typically parameterized using animal tracking data, and advances in animal tracking technology have led to increasing levels of autocorrelation between locations in such data sets. Because RSFs assume that data are independent and identically distributed, such autocorrelation can cause misleadingly narrow confidence intervals and biased parameter estimates. Data thinning, generalized estimating equations and step selection functions (SSFs) have been suggested as techniques for mitigating the statistical problems posed by autocorrelation, but these approaches have notable limitations that include statistical inefficiency, unclear or arbitrary targets for adequate levels of statistical independence, constraints in input data and (in the case of SSFs) scale‐dependent inference. To remedy these problems, we introduce a method for likelihood weighting of animal locations to mitigate the negative consequences of autocorrelation on RSFs. In this study, we demonstrate that this method weights each observed location in an animal's movement track according to its level of non‐independence, expanding confidence intervals and reducing bias that can arise when there are missing data in the movement track. Ecologists and conservation biologists can use this method to improve the quality of inferences derived from RSFs. We also provide a complete, annotated analytical workflow to help new users apply our method to their own animal tracking data using the ctmm R package.

Confronting School Resegregation.

School segregation is 1 of the mechanisms through which structural racism undermines the health of Black children in the United States.1,2 In this issue of Pediatrics, Wang et al3 examined how exposure to school segregation was associated with health outcomes among Black children represented in the longitudinal Panel Study of Income Dynamics (PSID). Using a quasi-experimental design, the authors focused on the impact of resegregation within school districts released by the courts from previous desegregation orders. The authors fit instrumental variable (IV) models to isolate this impact of school resegregation from the health consequences of residential segregation and other confounding factors. Based on the IV analysis, the authors found that school resegregation attributable to the lifting of desegregation orders was associated with increases in behavioral problems and drinking alcohol among Black youth. Yet, overall levels of school segregation were not associated with these or other health outcomes in the analysis which used conventional multivariable adjustment. As such, the results of Wang et al3 underscore the need to better understand how and when school segregation most threatens the health of Black children.Resegregation of US schools is linked to a series of court decisions beginning in the 1970s, which released school districts from responsibility to undo the persistent segregation of Black and White children.4 Releases from desegregation orders accelerated in the 1990s, providing the basis for the IV analysis used by Wang et al.3 This history of school resegregation reveals the extent to which school resegregation is politically driven, and not only reflective of residential segregation.5 Indeed, more recent cohorts of Black children have experienced increasing exposure to school segregation, even as levels of Black-White residential segregation have trended downward.6,7 Several aspects of the school system have been implicated in the resurgence of school segregation, including enrollment in private and charter schools and growing disparities between adjoining school districts.8,9 However, some research suggests that the most consequential actions for school resegregation are those of individual school districts.9,10 For example, in a case study of 3 school districts in areas with high residential segregation, the 2 districts which actively considered diversity in student assignment policies achieved meaningfully lower levels of school segregation.10The importance of local context for school resegregation presents a challenge to analyzing health outcomes of resegregation in national data sets such as the PSID. Despite its nationally representative design, the PSID was not intended to generate estimates for small geographies, and in the current study, a sample of only 1248 Black children was available to represent ∼1000 school districts that had been under desegregation orders in 1991.3 This limitation, coupled with the statistical inefficiency of IV methods, suggests that the current study likely underestimated the impact of school resegregation on health. Furthermore, Black children’s exposure to school segregation is highest in the elementary grades,6 whereas many of the health outcomes available in the PSID are more applicable to older children and adolescents (eg, smoking and drinking). As such, measures of health status, health care access, and development tailored to evaluating younger children might have revealed more fully the adverse effects of school resegregation.Despite these limitations, the work of Wang et al3 is a timely reminder that the harms of school segregation persist and extend beyond the schoolhouse walls. During the ongoing coronavirus disease 2019 pandemic, the rise of virtual learning has contributed to further segregation of the educational experience, even among children nominally enrolled in the same schools and classes. Notwithstanding larger structural forces, such as residential segregation, and the role of individual families’ beliefs and choices, recent research suggests that action at the level of schools and districts can have critical implications for exacerbating or reversing trends in school resegregation.8,10 Building on the current study, future work leveraging local health data would help reveal how advocacy at this level can undo the harm school segregation has caused and is continuing to cause.

Decoding with confidence: Statistical control on decoder maps

In brain imaging, decoding is widely used to infer relationships between brain and cognition, or to craft brain-imaging biomarkers of pathologies. Yet, standard decoding procedures do not come with statistical guarantees, and thus do not give confidence bounds to interpret the pattern maps that they produce. Indeed, in whole-brain decoding settings, the number of explanatory variables is much greater than the number of samples, hence classical statistical inference methodology cannot be applied. Specifically, the standard practice that consists in thresholding decoding maps is not a correct inference procedure. We contribute a new statistical-testing framework for this type of inference. To overcome the statistical inefficiency of voxel-level control, we generalize the Family Wise Error Rate (FWER) to account for a spatial tolerance δ, introducing the δ-Family Wise Error Rate (δ-FWER). Then, we present a decoding procedure that can control the δ-FWER: the Ensemble of Clustered Desparsified Lasso (EnCluDL), a procedure for multivariate statistical inference on high-dimensional structured data. We evaluate the statistical properties of EnCluDL with a thorough empirical study, along with three alternative procedures including decoder map thresholding. We show that EnCluDL exhibits the best recovery properties while ensuring the expected statistical control.

Value of NMR relaxation parameters of diamagnetic molecules for quantum information processing: optimizing the coherent phase

Quantum computing is the science that studies the applications of quantum mechanics in computer science. Thus, a quantum computer is an entity which can store quantum information, which in turn is an innovative procedure that creates conditions to allow the implementation of quantum algorithms and simulations, wherein the decoherence is a key problem. Considering the unique qubits, the time of the coherence phase can be well measured by the time of transversal relaxation (T2). In this line, we study the PPN derivatives, PPN-F, PPN-ethyl and PPN-NH2, as candidates to carry out quantum information. Therefore, the structures were optimized at the B3LYP/6-311G(d,p) level; after the optimization, molecular dynamics (MD) simulations were performed and the structures obtained were selected by statistical inefficiency method to then obtain the relaxation parameters (T1 and T2). The modification in the PPN molecule increases the transverse relaxation rate phosphorus nuclear spins at most five times because of the hydrogen bonds. From our findings, the studied molecules showed excellent candidates for the processing of quantum information; however, it is important to rule out the PPN-NH2 molecule that has more significant values of T1 and T2, and thus, this compound stands out from the others studied.

Evaluating Technical Efficiency of Stock Performance using Copula-Based Stochastic Frontier Analysis Approach

This study estimates the technical efficiency for 14 selected Malaysian trading and services companies for 2017 and 2018 using the Cobb-Douglas with the Stochastic Frontier production model. In standard stochastic frontier analysis (SSFA), the two-sided error, statistical noise, and one-sided error, inefficiency term are assumed to be independent. The assumption for statistical noise is normally distributed and the inefficiency term is half-normal distributed. In this paper, the copula model is applied to capture the joint distribution of these two error components. The copula model is an alternative method that is able to account for the joint multivariate distribution. Copula functions can be used to capture rank correlation and tail dependence between two error components, thus making the SFA more flexible. Seven copula models from the Archimedean copula family are considered in this study including the copulas with the trigonometric and hyperbolic generator. This study further compares the technical efficiency yielded by copulas with standard SFA, DEA-CCR and DEA-BCC models. The results raise the question of the reliability of SFA based on the standard model. Since the dependence error between statistical noise and inefficiency error cannot be ignored, the copula-based in SFA models can be considered as an alternative suitable tool for measuring efficiency performance.

Geometric ergodicity of a more efficient conditional Metropolis-Hastings algorithm

Despite its extensive application in practice, the Metropolis-Hastings sampler can suffer from slow mixing and, in turn, statistical inefficiency. We introduce a modification to the Metropolis-Hastings algorithm that, under specified conditions, encourages more efficient movement on general state spaces while preserving the overall quality of convergence, geometric ergodicity in particular. We illustrate the modified algorithm and its properties for the Metropolis-Hastings algorithm for a toy univariate Normal model and for the Gibbs sampling algorithm in a toy bivariate Normal model and a Bayesian dynamic spatiotemporal model.

On Restraints in End-Point Protein-Ligand Binding Free Energy Calculations.

The impact of harmonic restraints on protein heavy atoms and ligand atoms on end-point free energy calculations is systematically characterized for 54 protein-ligand complexes. We observe that stronger restraints reduce the equilibration time and statistical inefficiency, suppress conformational sampling, influence correlation with experiment, and monotonically decrease the estimated loss of entropy upon binding, leading to stronger estimated binding free energies in most systems. A statistical estimator that reweights for the biasing potential and includes data prior to the estimated equilibration time has the highest correlation with experiment. A spring constant of 20 cal mol-1 Å-2 maintains a near-native energy landscape and suppresses artifactual energy minima while minimally limiting thermal fluctuations about the crystal structure. © 2019 Wiley Periodicals, Inc.

On Restraints in End-Point Protein-Ligand Binding Free Energy Calculations

The impact of harmonic restraints on protein heavy atoms and ligand atoms on end-point free energy calculations is systematically characterized for 54 protein-ligand complexes. We observe that stronger restraints reduce the equilibration time and statistical inefficiency, suppress conformational sampling, influence correlation with experiment, and monotonically decrease the estimated loss of entropy upon binding, leading to stronger estimated binding free energies in most systems. A statistical estimator that reweights for the biasing potential and includes data prior to the estimated equilibration time has the highest correlation with experiment. A spring constant of 20 cal mol-1 A-2 maintains a near-native energy landscape and suppresses artifactual energy minima while minimally limiting thermal fluctuations about the crystal structure. © 2019 Wiley Periodicals, Inc.

Sparse Partially Collapsed MCMC for Parallel Inference in Topic Models

ABSTRACTTopic models, and more specifically the class of latent Dirichlet allocation (LDA), are widely used for probabilistic modeling of text. Markov chain Monte Carlo (MCMC) sampling from the posterior distribution is typically performed using a collapsed Gibbs sampler. We propose a parallel sparse partially collapsed Gibbs sampler and compare its speed and efficiency to state-of-the-art samplers for topic models on five well-known text corpora of differing sizes and properties. In particular, we propose and compare two different strategies for sampling the parameter block with latent topic indicators. The experiments show that the increase in statistical inefficiency from only partial collapsing is smaller than commonly assumed, and can be more than compensated by the speedup from parallelization and sparsity on larger corpora. We also prove that the partially collapsed samplers scale well with the size of the corpus. The proposed algorithm is fast, efficient, exact, and can be used in more modeling situations than the ordinary collapsed sampler. Supplementary materials for this article are available online.

SOFTWARE EFFORT ESTIMATION FRAMEWORK TO IMPROVE ORGANIZATION PRODUCTIVITY USING EMOTION RECOGNITION OF SOFTWARE ENGINEERS IN SPONTANEOUS SPEECH

Productivity is a very important part of any organisation in general and software industry in particular. Now a day's Software Effort estimation is a challenging task. Both Effort and Productivity are inter-related to each other. This can be achieved from the employee's of the organization. Every organisation requires emotionally stable employees in their firm for seamless and progressive working. Of course, in other industries this may be achieved without man power. But, software project development is labour intensive activity. Each line of code should be delivered from software engineer. Tools and techniques may helpful and act as aid or supplementary. Whatever be the reason software industry has been suffering with success rate. Software industry is facing lot of problems in delivering the project on time and within the estimated budget limit. If we want to estimate the required effort of the project it is significant to know the emotional state of the team member. The responsibility of ensuring emotional contentment falls on the human resource department and the department can deploy a series of systems to carry out its survey. This analysis can be done using a variety of tools, one such, is through study of emotion recognition. The data needed for this is readily available and collectable and can be an excellent source for the feedback systems. The challenge of recognition of emotion in speech is convoluted primarily due to the noisy recording condition, the variations in sentiment in sample space and exhibition of multiple emotions in a single sentence. The ambiguity in the labels of training set also increases the complexity of problem addressed. The existing models using probabilistic models have dominated the study but present a flaw in scalability due to statistical inefficiency. The problem of sentiment prediction in spontaneous speech can thus be addressed using a hybrid system comprising of a Convolution Neural Network and Hidden Markov Model.

Probing thermal and solvent effects on hyperfine interactions and spin relaxation rate of δ-FeOOH(1 0 0) and [MnH3buea(OH)]2−: Toward new MRI probes

Cancer is a global epidemic, which significantly affects all ages and socio-economic groups and one reason is the great difficulty of the initial diagnostic phase. The Magnetic Resonance Imaging (MRI), through effective contrast agents, has greatly helped in the diagnosis at the initial stage. Recently, superparamagnetic iron oxide nanoparticles and complexes with Mn2+ have received great attention due to their applications as contrast agents for Magnetic Resonance Imaging (MRI). Those materials can shorten the T2 and T2∗ transverse relaxation times. Thus, in this work, the face 100 of the δ-FeOOH and the complex [MnH3buea(OH)]2− were studied with explicit water molecules in order to obtain the 1H and 17O hyperfine coupling constants (HFCCs). Molecular dynamics (MD) simulations were performed using the ReaxFF program for subsequent statistical inefficiency calculations. Thus, the structures from the MD simulation were selected for HFCC calculations. The theoretical results suggest that the δ-FeOOH and [MnH3buea(OH)]2− considerably increase the 1H and 17O hyperfine coupling constants of the water molecules. In addition, δ-FeOOH is sensitive to 1H and 17O HFCCs parameters, however, in the complex [MnH3buea(OH)]2−, 17O is much more sensitive than 1H in relation to the HFCC parameters. Our findings point out δ-FeOOH as a promising alternative to conventional contrast agents.

Dynamics, NMR parameters and hyperfine coupling constants of the Fe3O4(1 0 0)–water interface: Implications for MRI probes

Magnetite is an iron oxide widely used as contrast agent in MRI, receiving considerable interest from nanoscience and nanotechnology. In this work, the face 100 of the magnetite structure was studied with water in order to obtain 1H hyperfine coupling constants (HFCCs). Molecular dynamics (MD) calculations were performed using the ReaxFF program and for statistical inefficiency, structures were selected for HFCC and NMR calculations. From our theoretical findings, the magnetite in solution considerably increases the 1H HFCC of water molecules. From our results, it is essential to incorporate the dynamics and solvent effects into NMR calculations of relaxation parameters.

The consumption-real exchange rate anomaly with extensive margins

This paper investigates a consumption-real exchange rate anomaly from the open macroeconomics literature known as the Backus-Smith puzzle. We both analytically and quantitatively examine how an expansion of trade along extensive margins can contribute to the puzzle's resolution. Our argument is based on 1) a wealth effect due to changes in the number of product varieties, 2) statistical inefficiency in measuring the number of product varieties, and 3) market incompleteness. Contrary to complete asset markets which, in general, feature overly strong risk sharing properties, changes in the number of product varieties under incomplete markets may produce a wealth effect under high trade elasticity. Since statistical agencies systematically fail to capture the welfare impact arising from that changes, data-consistent terms of trade and real exchange rates tend to appreciate due to this positive wealth effect. This provides a realistic correlation between data-consistent real exchange rates and consumption.

Stochastic FDH/DEA estimators for frontier analysis

In this paper we extend the work of Simar (J Product Ananl 28:183–201, 2007) introducing noise in nonparametric frontier models. We develop an approach that synthesizes the best features of the two main methods in the estimation of production efficiency. Specifically, our approach first allows for statistical noise, similar to Stochastic frontier analysis (even in a more flexible way), and second, it allows modelling multiple-inputs-multiple-outputs technologies without imposing parametric assumptions on production relationship, similar to what is done in non-parametric methods, like Data Envelopment Analysis (DEA), Free Disposal Hull (FDH), etc.... The methodology is based on the theory of local maximum likelihood estimation and extends recent works of Kumbhakar et al. (J Econom 137(1):1–27, 2007) and Park et al. (J Econom 146:185–198, 2008). Our method is suitable for modelling and estimation of the marginal effects onto inefficiency level jointly with estimation of marginal effects of input. The approach is robust to heteroskedastic cases and to various (unknown) distributions of statistical noise and inefficiency, despite assuming simple anchorage models. The method also improves DEA/FDH estimators, by allowing them to be quite robust to statistical noise and especially to outliers, which were the main problems of the original DEA/FDH estimators. The procedure shows great performance for various simulated cases and is also illustrated for some real data sets. Even in the single-output case, our simulated examples show that our stochastic DEA/FDH improves the Kumbhakar et al. (J Econom 137(1):1–27, 2007) method, by making the resulting frontier smoother, monotonic and, if we wish, concave.

On the application of accelerated molecular dynamics to liquid water simulations.

Our group recently proposed a robust bias potential function that can be used in an efficient all-atom accelerated molecular dynamics (MD) approach to simulate the transition of high energy barriers without any advance knowledge of the potential-energy landscape. The main idea is to modify the potential-energy surface by adding a bias, or boost, potential in regions close to the local minima, such that all transitions rates are increased. By applying the accelerated MD simulation method to liquid water, we observed that this new simulation technique accelerates the molecular motion without losing its microscopic structure and equilibrium properties. Our results showed that the application of a small boost energy on the potential-energy surface significantly reduces the statistical inefficiency of the simulation while keeping all the other calculated properties unchanged. On the other hand, although aggressive acceleration of the dynamics simulation increases the self-diffusion coefficient of water molecules greatly and dramatically reduces the correlation time of the simulation, configurations representative of the true structure of liquid water are poorly sampled. Our results also showed the strength and robustness of this simulation technique, which confirm this approach as a very useful and promising tool to extend the time scale of the all-atom simulations of biological system with explicit solvent models. However, we should keep in mind that there is a compromise between the strength of the boost applied in the simulation and the reproduction of the ensemble average properties.

Community-Based Tai Chi and Its Effect on Injurious Falls, Balance, Gait, and Fear of Falling in Older People

It is important to determine the effect of adherence to a tai chi program on falls and related functional outcomes in older people. This study examined the effect of a community-based tai chi program on injurious falls, balance, gait, and fear of falling among people aged 65 years and older in Taiwan. In 6 rural villages in Taichung County, 1,200 subjects participated in the initial assessment. During a 1-year intervention period, all study villages were provided with education on fall prevention. Two villages had been provided tai chi exercise (n=472 participants or "tai chi villagers"), and 4 villages served as control villages (n=728 participants or "control villagers"). Injurious falls were ascertained by telephone interviews every 3 months over a 2-year study period; additionally, balance, gait, and fear of falling were assessed in 2 follow-up assessments. Eighty-eight subjects, 83 from the tai chi villages and 5 from the control villages, participated and practiced in the tai chi program (the group labeled "tai chi practitioners"). After the tai chi program, injurious falls among the control villagers significantly declined by 44% (adjusted rate ratio [RR]=0.56; 95% confidence interval [CI]=0.36-0.92). Compared with the results for the control villagers, the decline was 31% greater (RR=0.69; 95% CI=0.30-1.56) among the tai chi villagers and 50% greater (RR=0.5; 95% CI=0.11-2.17) among the tai chi practitioners; the results did not reach statistical significance. Furthermore, compared with the scores for the control villagers, the scores for the tai chi practitioners increased by 1.8 points (95% CI=0.2-3.4) on the Tinetti Balance Scale and increased by 0.9 point (95% CI=0.1-1.8) on the Tinetti Gait Scale. No significant changes in the fear of falling were detected among the tai chi practitioners, tai chi villagers, and control villagers. Tai chi can prevent a decline in functional balance and gait among older people. However, the reduction in injurious falls attained with tai chi did not reach statistical significance; the statistical inefficiency may have resulted partly from the large decline in injurious falls in control villagers. Finally, the unexpected effect of educational intervention on reducing injurious falls in different settings needs to be further examined.