Two-step Estimation Research Articles

CBREN: Convolutional Neural Networks for Constant Bit Rate Video Quality Enhancement

Constant bit rate (CBR) videos are widely used in streaming playback applications. However, the image quality of the CBR video is often unstable, especially for scenes with large motion. To this end, we design a new model to represent the distortion of High Efficiency Video Coding (HEVC) constant bit rate video, and propose a neural network for a constant bit rate video quality enhancement (CBREN). We propose a dual-domain restoration module (DRM) to jointly learn the prior knowledge in the pixel domain and the frequency domain. To address the degradation resulting from compression, we propose a two-step quantization degradation estimation strategy. The Inverse DCT (IDCT) Translation Unit (ITU) is used to constrain the quantization table of the constant bit rate video to a suitable range, and the Dynamic Alpha Unit (DAU) is used to fine-tune the quantization table according to the content of each frame. In order to effectively reduce the block distortion of different sizes produced in the compression process, we adopt a multi-scale network. Extensive experiments show that our approach can greatly enhance the quality of CBR compressed video. Moreover, our method can also be applied to constant quantization parameter (CQP) video enhancement tasks, and is certainly superior to existing methods.

Interdependent capital structure choices and the macroeconomy

This study shows that capital structure choices of US corporations are interdependent across time. We follow a two-step estimation approach. First, using a large cross-section of firms we estimate year-by-year average capital structure choices, i.e., the average firm’s percentage of new funding that is secured through debt, its term composition, and the percentage of new equity represented by retained earnings. Second, these time series are included in a Factor Augmented Vector Autoregressive model in which three factors representing real economic activity, expected future funding conditions, and prices, are included. We test for the interdependence between optimal capital structure decisions and for the influence exerted by macroeconomic conditions on these decisions. Results show there is a hierarchical order in which firms make capital structure decisions. They first decide on the share of debt out of total new funding they will hire. Conditional on this they decide on the term of their debt and on their earnings retention policy. Of outmost importance, macroeconomic factors are key for making capital structure decisions.

Determinants of Credit Risk in Ethiopian Banking Industry: Does Political Stability Matter?

The continued political instability in Ethiopia’s Tigray Region has made it tough for bankers to recover loan repayments, triggering banks’ nonperforming assets to rise. Beyond the bank industry-specific and macroeconomic determinants, the article aims to understand the relationship between political stability and credit risk. The article focuses on 16 Ethiopian banks to understand the mentioned relationship in developing countries. From 2010 to 2020, data are obtained from the World Governance Indicators, World Development Indicators database and Ethiopia’s National Bank. Based on a two-step dynamic panel estimator, the results suggest that political stability is negatively correlated with credit risk, while political instability does have a positive impact on bank credit risk. Credit risk was inversely related to increased operating efficiency, diversified earnings and national currency appreciation. In contrast, greater bank size, rising inflation and economic expansion all contribute to a high level of credit risk. The loan-to-asset ratio and profitability were shown to have a minor influence on nonperforming loans (NPLs). Thus, this study provides a better understanding of the effects of political stability on the factors that determine levels of NPLs in developing economies.

LiDAR Odometry by Deep Learning-Based Feature Points with Two-Step Pose Estimation

An accurate ego-motion estimation solution is vital for autonomous vehicles. LiDAR is widely adopted in self-driving systems to obtain depth information directly and eliminate the influence of changing illumination in the environment. In LiDAR odometry, the lack of descriptions of feature points as well as the failure of the assumption of uniform motion may cause mismatches or dilution of precision in navigation. In this study, a method to perform LiDAR odometry utilizing a bird’s eye view of LiDAR data combined with a deep learning-based feature point is proposed. Orthographic projection is applied to generate a bird’s eye view image of a 3D point cloud. Thereafter, an R2D2 neural network is employed to extract keypoints and compute their descriptors. Based on those keypoints and descriptors, a two-step matching and pose estimation is designed to keep these feature points tracked over a long distance with a lower mismatch ratio compared to the conventional strategy. In the experiment, the evaluation of the proposed algorithm on the KITTI training dataset demonstrates that the proposed LiDAR odometry can provide more accurate trajectories compared with the handcrafted feature-based SLAM (Simultaneous Localization and Mapping) algorithm. In detail, a comparison of the handcrafted descriptors is demonstrated. The difference between the RANSAC (Random Sample Consensus) algorithm and the two-step pose estimation is also demonstrated experimentally. In addition, the data collected by Velodyne VLP-16 is also evaluated by the proposed solution. The low-drift positioning RMSE (Root Mean Square Error) of 4.70 m from approximately 5 km mileage shown in the result indicates that the proposed algorithm has generalization performance on low-resolution LiDAR.

Integrated Calibration of Strap-Down Geomagnetic Vector Measurement System

Strap-down geomagnetic vector measurement system usually consists of three-axis magnetometer and inertial units, which is an effective device for measuring geomagnetic vector. However, its measurement accuracy is limited by disturbance of ferromagnetic material, measurement error of magnetometer and error caused by misalignment. Therefore, an integrated calibration method for strap-down geomagnetic vector measurement system is proposed. Firstly, the integrated error model is established, and a system with linear equations constraint is constructed by the measurements of inertial units, three-axis magnetometer and proton magnetometer; secondly, calibration parameters can be estimated by two-step least squares estimation (LSE); finally, geomagnetic vector measurement can be calibrated with these parameters in real time. Simulation results show that the relative errors between estimation and true parameters are less than 3%. Experimental results demonstrate that the errors of strap-down geomagnetic vector measurement system are suppressed effectively after integrated calibration, where North, Vertical and East component measurement errors are reduced from 809.2 nT, 300.0 nT, 776.8 nT to 14.8 nT (1.83% of uncalibrated value), 11.9 nT (3.97%), 32.5 nT (4.18%) respectively, and measurement accuracy of declination and inclination are achieved to 0.1°. Furthermore, the calibration error and time are least compared with traditional methods.

Machine Learning-Based Gray-Level Co-Occurrence Matrix (GLCM) Models for Predicting the Depth of Myometrial Invasion in Patients with Stage I Endometrial Cancer.

PurposeDeep myometrial invasion (DMI) is an independent high-risk factor for lymph node metastasis and a prognostic risk factor in early-stage endometrial cancer (EC-I) patients. Thus, we developed a machine learning (ML) assistant model, which can accurately help define the surgical area.Methods348 consecutive EC-I patients with the pathological diagnosis were recruited in the tertiary medical centre between January 1, 2012, and October 31, 2021. Five ML-assisted models were developed using two-step estimation methods from the candidate gray level co-occurrence matrix (GLCM). Receiver operating characteristic curve (ROC), decision curve analysis (DCA), and clinical impact curve (CIC) were prepared to evaluate the robustness and clinical practicality of each model.ResultsOur analysis identified several significant differences between the stage IA and IB groups. The top seven-candidate factors included correlation all direction offset1, correlation angle0 offset1, correlation angle45 offset1, correlation angle90 offset1, ID moment all direction offset1, ID moment angle0 offset1, and ID moment angle45 offset1. The areas under the ROC curve (AUCs) of the random forest classifier (RFC) model, support vector machine (SVM), eXtreme gradient boosting (XGBoost), artificial neural network (ANN), and decision tree (DT) ranged from 0.765 to 0.877 in the training set and from 0.716 to 0.862 in the testing set, respectively. Among the five machine algorithms, RFC obtained the optimal prediction efficiency using correlation angle0 offset1, correlation angle45 offset1, correlation angle90 offset1, correlation all direction offset1, ID moment angle0 offset1, and ID moment angle45 offset1, and ID moment angle90 offset1, respectively.ConclusionOur ML-based prediction model combined with GLCM parameters assessed the risk of DMI in EC-I patients, especially RFC, which helped distinguish stage IA and IB EC patients. This new predictive model based on supervised learning can be used to establish personalized treatment strategies.

Manly transformation in quantile regression: A comparison of two transformation parameter estimators

This study implements the Manly transformations for normalization of variables in quantile regression analysis.The transformation parameter was estimated using two different methods namely; the maximum likelihood estimation (MLE) method and the two-step estimation method by Chamberlain and Buchinsky(CBTS).The transformation parameters obtained using the two different methods were used for the Manly transformation of data with outliers and data without outliers. The methods were applied to a quantile regression analysis at different quantiles (0.25, 0.50, 0.75, 0.95). Based on our findings, for data without outliers, the 25th quantile model was seen to be the best fit model compared to the other quantiles for the CBTS method with AIC=-43.46279, BIC=20.75212 and MSE=0.70956, while for the MLE the 50th quantile model was seen to be the best fit model with AIC=-348.3657, BIC=20.13548, and MSE=0.00864. Considering data with outliers the 25th quantile model was still seen to be the best fit model compared to the other quantiles for the CBTS method with AIC=-48.5671, BIC=21.8321 and MSE=0.92341, while for the MLE the 50th quantile model was still seen to be the best fit model with AIC=988.6763, BIC=710.09, and MSE=690.7965. Comparison of both methods for data without outliers the study concludes that the estimation of the transformation parameter using the MLE produced better results with lower AIC, BIC and MSE at all quantiles and for data with outliers the study concludes that the estimation of the transformation parameter using CBTS produced better results with lower AIC, BIC and MSE results as is shown in table (3.5) and table (3.6) respectively.

Two-step series estimation and specification testing of (partially) linear models with generated regressors

This paper studies three semiparametric models that are useful and frequently encountered in applied econometric work—a linear and two partially linear specifications with generated regressors, i.e., the regressors that are unobserved, but can be nonparametrically estimated from the data. Our framework allows for generated regressors to appear in linear or nonlinear components of partially linear models. We propose two-step series estimators for the finite-dimensional parameters, establish their -consistency (with sample size n) and asymptotic normality, and provide the asymptotic variance formulae that take into account the estimation error of generated regressors. Moreover, we develop a nonparametric specification test for the models considered. Numerical performances of the proposed estimators and test via simulation experiments and an empirical application illustrate the utility of our approach.

Climate Adaptive Response Estimation: Short and long run impacts of climate change on residential electricity and natural gas consumption

This paper proposes a simple two-step estimation method (Climate Adaptive Response Estimation - CARE) to estimate sectoral climate damage functions, which account for long-run adaptation. The paper applies this method in the context of residential electricity and natural gas demand for the world’s fifth largest economy — California. The advantage of the proposed method is that it only requires detailed information on intensive margin behavior, yet does not require explicit knowledge of the extensive margin response (e.g., technology adoption). Using almost two billion energy bills, we estimate spatially highly disaggregated intensive margin temperature response functions using daily variation in weather. In a second step, we explain variation in the slopes of the dose response functions across space as a function of summer climate. Using 18 climate models, we simulate future demand by letting households vary consumption along the intensive and extensive margins. We show that failing to account for extensive margin adjustment in electricity demand leads to a significant underestimate of the future impacts on electricity consumption. We further show that reductions in natural gas demand more than offset any climate-driven increases in electricity consumption in this context.

Informality and Firm Performance in Myanmar

Using a unique panel survey of enterprises, we examine the relationship between four categories of formalization and firm productivity. We carry out one- and two-step productivity estimations whose robustness we check with matching and doubly robust estimators. The only formalization category that appears to be significantly associated with productivity is tax formalization, i.e. a firm’s decision to pay taxes. This positive association only holds for firms that were already more productive and bigger before formalizing than other informal firms. The reason for the insignificance of the remaining three categories is likely to be the insignificant association between formalization and potential benefits of formalization, such as more access to credit, employees, and investments. High taxes and fees linked to formalization seem to outweigh the few to non-existent intermediate benefits of formalization.

Estimation of a partially linear seemingly unrelated regressions model: application to a translog cost system

This article studies a partially linear seemingly unrelated regressions (SUR) model to estimate a translog cost system that consists of a partially linear translog cost function and input share equations. The parametric component is estimated via a simple two-step feasible SUR estimation procedure. We show that the resulting estimator achieves root-n convergence and is asymptotically normal. The nonparametric component is estimated with a nonparametric SUR estimator based on the Cholesky decomposition. We show that this estimator is consistent, asymptotically normal, and more efficient relative to the ones that ignore cross-equation correlation. We emphasize the importance and implication of the choice of square root of the covariance matrix by comparing the Cholesky and Spectral decompositions. A model specification test for parametric functional form is proposed. An Italian banking data set is used to estimate the translog cost system. Results show that marginal effects of risks on cost of production are heterogeneous but increase with risk levels.

Renewable energy and CO2 emissions intensity in the top carbon intense countries

This study employed panel data techniques to explore the relationship between renewable energy and CO2 emissions intensity in the top carbon intense countries over the period 2000–2015. The results based on fixed effects regression and two-step GMM estimator show that renewable energy has significant negative effect on CO2 emissions. The estimates suggest that a percentage increase in renewable energy consumption leads to a decrease of 0.98% in CO2 emissions. GDP per capita has negative impact on carbon emissions. A 1% increase in GDP per capita leads to 0.21% decrease in CO2 emissions. We also find that increase in the share of women in parliament has positive effect on environmental quality.

Robust Estimation of Probit Models with Endogeneity

Probit models with endogenous regressors are commonly used models in economics and other social sciences. Yet, the robustness properties of parametric estimators in these models have not been formally studied. The influence functions of the endogenous probit model’s classical estimators (the maximum likelihood and the two-step estimator) are derived and their non-robustness to small but harmful deviations from distributional assumptions is proven. A procedure to obtain a robust alternative estimator is proposed, its asymptotic normality is proven and its asymptotic variance is provided. A simple robust test for endogeneity is also constructed. The performance of the robust and classical estimators is compared in Monte Carlo simulations with different types of contamination scenarios. The use of the robust estimator is illustrated in several empirical applications.

Machine Learning-Assisted Ensemble Analysis for the Prediction of Acute Pancreatitis with Acute Kidney Injury.

PurposeAcute kidney injury (AKI) is a frequent complication of severe acute pancreatitis (AP) and carries a very poor prognosis. The present study aimed to construct a model capable of accurately identifying those patients at high risk of harboring occult acute kidney injury (AKI) characteristics.Patients and MethodsWe retrospectively recruited a total of 424 consecutive patients at the Gezhouba central hospital of Sinopharm and Xianning central hospital between January 1, 2016, and October 30, 2021. ML-assisted models were developed from candidate clinical features using two-step estimation methods. The receiver operating characteristic curve (ROC), decision curve analysis (DCA), and clinical impact curve (CIC) were performed to evaluate the robustness and clinical practicability of each model.ResultsFinally, a total of 30 candidate variables were included, and the AKI prediction model was established by an ML-based algorithm. The areas under the ROC curve (AUCs) of the random forest classifier (RFC) model, support vector machine (SVM), eXtreme gradient boosting (XGBoost), artificial neural network (ANN), and decision tree (DT) ranged from 0.725 (95% CI 0.223–1.227) to 0.902 (95% CI 0.400–1.403). Among them, RFC obtained the optimal prediction efficiency via adding inflammatory factors, which are serum creatinine (Scr), C-reactive protein (CRP), platelet-to-lymphocyte ratio (PLR), neutrophil-to-lymphocyte ratio (NLR), neutrophil-to-albumin ratio (NAR), and CysC, respectively.ConclusionWe successfully developed ML-based prediction models for AKI, particularly the RFC, which can improve the prediction of AKI in patients with AP. The practicality of prediction and early detection may be greatly beneficial to risk stratification and management decisions.

Impact of Foreign Direct Investment on Economic Growth in the North Provinces of Vietnam in the Quality of Institutions Government

The impact of foreign direct investment (FDI) on economic growth has long been of interest to researchers and policy makers. However, the results of this effect are still controversial; some studies suggest that FDI has the ability to promote economic growth. Some other studies suggest that FDI has a negative effect on economic growth. In this study, the authors examine the impact of FDI on economic growth in the Northern provinces of Vietnam in terms of the quality of institutional governance in these provinces in the period 2011-2020. By the two-step SGMM estimation method, the research results show that FDI has a positive impact on economic growth. At the same time, institutional quality is one of the causes affecting the relationship between FDI and economic growth. On the basis of research results, the author provides policy implications for the Northern provinces of Vietnam.

Intra-Household Gender Wage Gap in China: Evidence from the China Health & Nutrition Survey

This paper analyzes the intra-household wage gap and wage determination of husbands and wives under the urban-rural dual economic structure in China. A two-step estimation procedure was used to determine the factors that are related to living in urban or rural areas by employing the probit model. Additionally, the effects of different factors affecting gender wage across regimes and between husbands and wives was examined using the seemingly unrelated regression estimation (SURE) model. It was found that the intra-household gender wage gap is larger in rural areas than in urban areas. At different percentiles, the gender wage gap is also different. There is a larger gender wage gap between low income earners than higher income earners. The largest gender wage gap is in the 10th percentile in rural areas with a wife–husband wage ratio of 67%. The wage determination pattern between spouses is different in urban and rural areas. Except for human capital impact, urban-rural dual economic structure and family factors, spousal factors also effect the intra-household wage gap.

Two-step calibration estimation under multiple auxiliary variables in survey sampling

The use of auxiliary information usually improves the precision of estimation of finite population total in survey sampling. In this paper, we put forward a two-step calibration estimator of finite population total when multiple auxiliary variables are available. We study the statistical properties of the proposed estimator, suggest a method to select its parameters and carry out a simulation study to assess its performance. Results show that the proposed estimator outperforms others in terms of precision for different population distributions. We then apply the estimator to a set of real data, showing that it improves the precision in the estimation of finite population total.

Correction

Correction

Projecting Mortality Rates Using a Markov Chain

We present a mortality model where future stochastic changes in population-wide mortality are driven by a finite-state hierarchical Markov chain. A baseline mortality in an initial ‘Alive’ state is calculated as the average logarithm of the observed mortality rates. There are several more ‘Alive’ states and a jump to the next ‘Alive’ state leads to a change (typically, an improvement) in mortality. In order to estimate the model parameters, we minimized a weighted average quadratic distance between the observed mortality rates and expected mortality rates. A two-step estimation procedure was used, and a closed-form solution for the optimal estimates of model parameters was derived in the first step, which means that the model could be parameterized very fast and efficiently. The model was then extended to allow for age effects whereby stochastic mortality improvements also depend on age. Forecasting relies on state space augmentation and an innovations state space time series model. We show that, in terms of forecasting, our model outperforms a naïve model of static mortality within a few years. The Markov approach also permits an exact computation of mortality indices, such as the complete expectation of life and annuity present values, which are key in the life insurance and pension industries.

Stochastic Event-Based Distributed Fusion Estimation Over Sensor Networks With Fading Channel

The problem of the stochastic event-based distributed fusion estimation for a class of Gaussian systems is investigated. Considering the deterministic event-triggers destroying the Gaussian property of system states, the stochastic event-triggered mechanisms (SETMs) are used, which also can relieve the network transmission burden. Under the stochastic transmission schedules, a two-step fusion estimation method is developed. The first step, with the consideration of channel fading, the local estimation of each sensor is proposed by using the measurements from itself and its neighbors. The second step, the fusion algorithm is designed to eliminate the disagreements among local estimations of each sensor. Finally, experiment is carried out to demonstrate the advantages of the proposed distributed fusion estimation.