Nonparametric Regression Splines Research Articles

ACCEPT 2·0: Recalibrating and externally validating the Acute COPD exacerbation prediction tool (ACCEPT).

SummaryBackgroundThe Acute Chronic Obstructive Pulmonary Disease (COPD) Exacerbation Prediction Tool (ACCEPT) was developed for individualised prediction of COPD exacerbations. ACCEPT was well calibrated overall and had a high discriminatory power, but overestimated risk among individuals without recent exacerbations. The objectives of this study were to 1) fine-tune ACCEPT to make better predictions for individuals with a negative exacerbation history, 2) develop more parsimonious models, and 3) externally validate the models in a new dataset.MethodsWe recalibrated ACCEPT using data from the Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points (ECLIPSE, a three-year observational study, 1,803 patients, 2,117 exacerbations) study by applying non-parametric regression splines to the predicted rates. We developed three reduced versions of ACCEPT by removing symptom score and/or baseline medications as predictors. We examined the discrimination, calibration, and net benefit of ACCEPT 2·0 in the placebo arm of the Towards a Revolution in COPD Health (TORCH, a three-year randomised clinical trial of inhaled therapies in COPD, 1,091 patients, 1,064 exacerbations) study. The primary outcome for prediction was the occurrence of ≥2 moderate or ≥1 severe exacerbation in the next 12 months; the secondary outcomes were prediction of the occurrence of any moderate/severe exacerbation or any severe exacerbation.FindingsACCEPT 2·0 had an area-under-the-curve (AUC) of 0·76 for predicting the primary outcome. Exacerbation history alone (current standard of care) had an AUC of 0·68. The model was well calibrated in patients with positive or negative exacerbation histories. Changes in AUC in reduced versions were minimal for the primary outcome as well as for predicting the occurrence of any moderate/severe exacerbations (ΔAUC<0·011), but more substantial for predicting the occurrence of any severe exacerbations (ΔAUC<0·020). All versions of ACCEPT 2·0 provided positive net benefit over the use of exacerbation history alone for some range of thresholds.InterpretationACCEPT 2·0 showed good calibration regardless of exacerbation history, and predicts exacerbation risk better than current standard of care for a range of thresholds. Future studies need to investigate the utility of exacerbation prediction in various subgroups of patients.FundingThis study was funded by a team grant from the Canadian Institutes of Health Research (PHT 178432).

Spline Nonparametric Regression to Analyze Factors Affecting Gender Empowerment Measure (GEM) in East Java

Gender is a multidimensional issue that's not limited to gender discrimination, but alsoincludes the economic, educational, and health aspects, which then become the focus of almost all the Sustainable Development Goals (SDGs). Evaluation of the development devoted to the perspective of the gender using several indicators, Gender Development Index (GDI) and Gender Empowerment Measure (GEM). GEM describes the role of women in the economic sphere and is measured by equality in political participation. GEM of East Java for 5 consecutive years (2014 – 2018) is lower than the average national GEM. This study aims to identify factors affecting GEM in East Java using nonparametric regression spline quadratic. The result ofthe regression model shows the factors affecting GEM East Java is the Labor Force Participation Rate(LFPR) population of women (), School Participation Rate(SPR) high school population of women (), Percentage of Population Female thatWorking in the formal sector (), sex ratio (), Percentage of Population Female that Working as members of People’s Representative Council (), Percentage of Population Female that working as Civil Servants (), and rate of women's income donations (). The model generates value of 93.74% and MAPE of 3.22%.This research contributes to the implementation of non-parametric spline regression in identifying various factors that influence social phenomena.

KOMPOS: Connecting Causal Knots in Large Nonlinear Time Series with Non-Parametric Regression Splines

Recovering causality from copious time series data beyond mere correlations has been an important contributing factor in numerous scientific fields. Most existing works assume linearity in the data that may not comply with many real-world scenarios. Moreover, it is usually not sufficient to solely infer the causal relationships. Identifying the correct time delay of cause-effect is extremely vital for further insight and effective policies in inter-disciplinary domains. To bridge this gap, we propose KOMPOS, a novel algorithmic framework that combines a powerful concept from causal discovery of additive noise models with graphical ones. We primarily build our structural causal model from multivariate adaptive regression splines with inherent additive local nonlinearities, which render the underlying causal structure more easily identifiable. In contrast to other methods, our approach is not restricted to Gaussian or non-Gaussian noise due to the non-parametric attribute of the regression method. We conduct extensive experiments on both synthetic and real-world datasets, demonstrating the superiority of the proposed algorithm over existing causal discovery methods, especially for the challenging cases of autocorrelated and non-stationary time series.

Penerapan Regresi Nonparametrik Spline dalam Memodelkan Faktor-faktor yang Mempengaruhi Indeks Pembangunan Manusia di Provinsi Sulawesi Selatan Tahun 2015

Abstrak. Human development index (HDI) is a measure used in monitoring and evaluating human development. Indicators used to measure HDI consists of three basic components of quality of life that is the life chances, knowledge and decent living standards. Several factors are thought to affect the HDI in the district/city in South Sulawesi province that labor force participation rates, the ratio of school pupils, overcrowding, health facilities, and the Gross Domestic Product (GDP). When HDI and these factors are plotted then shows the pattern of data that is not to follow a certain pattern, so that the data can be applied to the nonparametric regression model spline truncated. Selection of the best model seen from the point of knots and the minimum value of GCV. Based on research, the value of the minimum GCV is at three knots point is equal to 5.33 Rated amounting to 80.29%.Keywords: Human development Index, GCV, Nonparametric Regression Spline, Knot Points

Modelling Return on Assets (ROA) using nonparametric regression spline truncated for longitudinal data

Nonparametric regression is one of the statistical methods used to see the relationship between response variables and predictor variables with unknown curves. The method often used in estimating the curve of nonparametric regression is the spline method. Spline is one of the nonparametric regression approaches that have very specific and very good statistical and visual interpretations. Spline truncated nonparametric regression is not only used to modelling data with cross-section type but can be used to modelling longitudinal data. Longitudinal data is data whose observations are repeated on each subject within a certain time. In this research, the modelling of Return On Assets (ROA) will be based on the factors that influence it with using 30 subjects of go public banks listed on the Indonesia stock exchange during 2012-2018 using nonparametric regression spline truncated for longitudinal data, with the weighting matrix using the variance-covariance matrix of errors. Based on the analysis data that has been done by using the weighting matrix of the error variance-covariance matrix and one-knot point with 14 Increments, the best model is obtained, namely using the weighting matrix based of the error variance-covariance matrix one-knot point with the smallest GCV value of 15.12 with MSE value of 0.739 and R 2 value of 85.56%.

Inflation Rate Modelling in Indonesia

The purposes of this research were to analyse: (i) Modelling the inflation rate in Indonesia with parametric regression. (ii) Modelling the inflation rate in Indonesia using non-parametric regression spline multivariable (iii) Determining the best model the inflation rate in Indonesia (iv) Explaining the relationship inflation model parametric and non-parametric regression spline multivariable. Based on the analysis using the two methods mentioned the coefficient of determination (R2) in parametric regression of 65.1% while non-parametric amounted to 99.39%. To begin with, the factor of money supply or money stock, crude oil prices and the rupiah exchange rate against the dollar is significant on the rate of inflation. The stability of inflation is essential to support sustainable economic development and improve people's welfare. In conclusion, unstable inflation will complicate business planning business activities, both in production and investment activities as well as in the pricing of goods and services produced.DOI: 10.15408/etk.v15i2.3260

A progressive approach for processing satellite data by operational research

Satellite data, together with spatial technologies, have a vital importance in earth sciences to continuously monitor natural and physical processes. However, images taken by earth-observing satellites are often associated with uncertainties due to atmospheric effects (i.e., absorption and scattering by atmospheric gases and aerosols). In this paper, a more adaptable approach for the removal of atmospheric effects from satellite data is introduced within an operational research perspective by utilizing nonparametric regression splines. Regional atmospheric correction models via multivariate adaptive regression splines (MARS) are applied on a set of satellite images for Alps and Turkey to calculate surface reflectance values. A classical radiative transfer based atmospheric correction method is likewise employed on the same data set. The results are compared in terms of relative differences with respect to surface reflectance data. MARS provides significant improvement in the order of 40 and 37 % for Alps and Turkey, respectively.

Increased consumption and vasodilatory effect of nitrite during exercise.

This study investigated the effects of aerobic-to-anaerobic exercise on nitrite stores in the human circulation and evaluated the effects of systemic nitrite infusion on aerobic and anaerobic exercise capacity and hemodynamics. Six healthy volunteers were randomized to receive sodium nitrite or saline for 70 min in two separate occasions in an exercise study. Subjects cycled on an upright electronically braked cycle ergometer 30 min into the infusion according to a ramp protocol designed to attain exhaustion in 10 min. They were allowed to recover for 30 min thereafter. The changes of whole blood nitrite concentrations over the 70-min study period were analyzed by pharmacokinetic modeling. Longitudinal measurements of hemodynamic and clinical variables were analyzed by fitting nonparametric regression spline models. During exercise, nitrite consumption/elimination rate was increased by ∼137%. Cardiac output (CO), mean arterial pressure (MAP), and pulmonary artery pressure (PAP) were increased, but smaller elevation of MAP and larger increases of CO and PAP were found during nitrite infusion compared with placebo control. The higher CO and lower MAP during nitrite infusion were likely attributed to vasodilation and a trend toward decrease in systemic vascular resistance. In contrast, there were no significant changes in mean pulmonary artery pressures and pulmonary vascular resistance. These findings, together with the increased consumption of nitrite and production of iron-nitrosyl-hemoglobin during exercise, support the notion of nitrite conversion to release NO resulting in systemic vasodilatation. However, at the dosing used in this protocol achieving micromolar plasma concentrations of nitrite, exercise capacity was not enhanced, as opposed to other reports using lower dosing.

Inversion of top of atmospheric reflectance values by conic multivariate adaptive regression splines

Spatial technologies offer high flexibility to handle substantial amount of spatial data and wide range of modelling capabilities. Remotely sensed data are the most significant data source used in spatial technologies. However, it is often associated with uncertainties due to atmospheric effects (i.e. absorption and scattering by atmospheric gases and aerosols). Methods based on rigorous treatment of radiative transfer models still have some drawbacks in the inversion of top of atmospheric reflectance values to surface reflectance values on large numbers of satellite images. In this paper, our aim is to represent a more flexible (adaptive) approach for the regional atmospheric correction by employing nonparametric regression splines within the frame of inverse problems and modern techniques of continuous optimization. To achieve this objective, atmospheric correction models obtained through conic multivariate adaptive regression splines, which is an alternative method to multivariate adaptive regression splines by constructing a penalized residual sum of squares as a Tikhonov regularization problem, are applied on a set of satellite images in order to convert the top of atmospheric reflectance values into surface reflectance values. The results are compared with the ones obtained by both multivariate adaptive regression splines and a radiative transfer-based method.

The crs Package: Nonparametric Regression Splines for Continuous and Categorical Predictors

A new package crs is introduced for computing nonparametric regression (and quan- tile) splines in the presence of both continuous and categorical predictors. B-splines are em- ployed in the regression model for the contin- uous predictors and kernel weighting is em- ployed for the categorical predictors. We also de- velop a simple R interface to NOMAD, which is a mixed integer optimization solver used to com- pute optimal regression spline solutions.

Association of γ-Glutamyltransferase and Risk of Cancer Incidence in Men: A Prospective Study

Although several epidemiologic studies have shown that gamma-glutamyltransferase (GGT) is independently associated with cardiovascular disease and all-cause mortality, its relationship with cancer incidence remains widely unexplored. In several experimental models, the ability of cellular GGT to modulate crucial redox-sensitive functions has been established, and it thus may play a role in tumor progression, as has been repeatedly suggested. We prospectively investigated the association between GGT and risk of overall and site-specific cancer incidence in a large population-based cohort of 79,279 healthy Austrian men with serial GGT measurements. Median follow-up was 12.5 years. Adjusted Cox proportional hazards models were calculated to evaluate GGT as an independent predictor for cancer incidence, and nonparametric regression splines were fitted to flexibly capture the dose-response relationship. Elevated GGT significantly increased overall cancer risk, showing a clear dose-response relationship (P for GGT log-unit increase < 0.0001; P for trend < 0.0001). In comparison with the reference GGT concentration (25 units/L), we found adjusted relative risks (95% confidence intervals) equalling 1.19 (1.15-1.22) for GGT concentrations of 60 units/L, 1.32 (1.28-1.36) for 100 units/L, 1.67 (1.60-1.75) for 200 units/L, and 2.30 (2.14-2.47) for 400 units/L. In cancer site-specific models, GGT was significantly associated with malignant neoplasms of digestive organs, the respiratory system/intrathoracic organs, and urinary organs (all P < 0.0001). Age of participants significantly modified the association of GGT and cancer risk (P < 0.001), revealing markedly stronger associations in participants ages </=65 years. Our findings, for the first time, show that elevated GGT is significantly associated with increased cancer risk in men.

Réduction de la variance dans les sondages en présence d'information auxiliarie: Une approache non paramétrique par splines de régression

The author considers the use of auxiliary information available at population level to improve the estimation of finite population totals. She introduces a new type of model-assisted estimator based on nonparametric regression splines. The estimator is a weighted linear combination of the study variable with weights calibrated to the B-splines known population totals. The author shows that the estimator is asymptotically design-unbiased and consistent under conditions which do not require the superpopulation model to be correct. She proposes a design-based variance approximation and shows that the anticipated variance is asymptotically equivalent to the Godambe-Joshi lower bound. She also shows through simulations that the estimator has good properties.

Estimating the distribution of lag in the effect of short-term exposures and interventions: adaptation of a non-parametric regression spline model.

Information on the distribution of lag duration between exposure or intervention and the subsequent changes in risk can help in assessing the impact of exposure, predicting cost-effectiveness of intervention, and understanding the underlying biological mechanisms. Previous approaches focused more on optimizing the strength of the exposure-disease association than on directly estimating lag duration. We propose an alternative approach applicable to the analysis of the lagged effects of binary exposure variables. The density function of the distribution of lags is estimated based on flexible modelling of changes in hazard ratio of exposed versus unexposed subjects. The methodology is evaluated in a simulation study and is applied to the Framingham data to investigate the lagged effect of smoking cessation on coronary heart disease risk.

Fast spline smoothing via spectral factorization concepts

When tuning the smoothness parameter of nonparametric regression splines, the evaluation of the so-called degrees of freedom is one of the most computer-intensive tasks. In the paper, a closed-form expression of the degrees of freedom is obtained for the case of cubic splines and equally spaced data when the number of data tends to infinity. State-space methods, Kalman filtering and spectral factorization techniques are used to prove that the asymptotic degrees of freedom are equal to the variance of a suitably defined stationary process. The closed-form expression opens the way to fast spline smoothing algorithms whose computational complexity is about one-half of standard methods (or even one-fourth under further approximations).

NON-PROPORTIONAL HAZARDS WITH APPLICATION TO KIDNEY TRANSPLANT DATA

The Cox proportional hazards (PH) model is the popular methodfor modelling censored survival data.Cox PH model is the proportionality of hazards in which the hazard ratiois linear in the covariates. However this assumption may not hold in somesurvival studies. Therefore, digerent non-parametric regression methods havebeen proposed to estimate the hazard ratio as a function of time when theproportionality of hazards can not be assumed.In this study a piecewisemodel and a non-parametric regression spline model have been considered forthe non-proportional hazards. The models have been illustrated with kidneytransplant data