Search results for "polynomial regression"
showing 10 items of 27 documents
Age-related dermal collagen changes during development, maturation and ageing - a morphometric and comparative study.
2014
The tissue organisation of dermal collagen is gaining importance as a contributing factor both in development and ageing, as well as in skin maturation processes. In this work we aim to study different representative parameters of this structural organisation in 45 human skin samples of assorted ages, by means of image analysis. The variation of these parameters on the basis of age was assessed using several regression models (linear, quadratic and cubic). The area occupied by collagen was significantly reduced as a function of age in the papillary dermis (R(2) = 0.437, P < 0.0001), as well as the thickness of the collagen bundles (R(2) = 0.461, P < 0.0001), following statistical models of …
Retrieval of coloured dissolved organic matter with machine learning methods
2017
The coloured dissolved organic matter (CDOM) concentration is the standard measure of humic substance in natural waters. CDOM measurements by remote sensing is calculated using the absorption coefficient (a) at a certain wavelength (e.g. 440nm). This paper presents a comparison of four machine learning methods for the retrieval of CDOM from remote sensing signals: regularized linear regression (RLR), random forest (RF), kernel ridge regression (KRR) and Gaussian process regression (GPR). Results are compared with the established polynomial regression algorithms. RLR is revealed as the simplest and most efficient method, followed closely by its nonlinear counterpart KRR.
The Norm-P Estimation of Location, Scale and Simple Linear Regression Parameters
1989
A new formulation of the exponential power distributions is used as general error model to describe long-tailed and short -tailed distributed errors. The proposed estimators of the location, scale and structure parameters of this general model and of the simple linear regression parameters when the response variable is affected by errors coming from the previous model should be used instead of robust estimators and against the practice of rejecting outlying observations. Two Monte Carlo simulations prove the good properties of these norm-p estimators.
Disposition-Content Congruency in Adolescents’ Alcohol-Related Social Media (Self-) Effects: The Role of the Five-Factor Model
2019
Objective: Accumulating evidence indicates that social networking sites play an increasingly important role in young people’s drinking behavior. The present study adds to this research by assessing the conditionality of the relationships between exposure to and self-sharing of alcohol-related content on social media and adolescents’ drinking behavior. Specifically, the moderating role of the five-factor model of personality is determined. Method: A cross-sectional survey study was conducted among 866 mid-adolescents (Msubsample = 14.85 years, SD = 0.71, 57.5% girls). Polynomial regression analysis with response surface modeling was used to test the interactions. Results: Exposure, but not s…
Monte Carlo calculation of dose rate distributions around the Walstam CDC.K-type137Cs sources
2001
Basic dosimetric data for the Walstam CDC.K-type low dose rate 137Cs sources in water have been calculated using Monte Carlo techniques. These sources, CDC.K1 -K3 and CDC.K4, are widely used in a range of applicators and moulds for the treatment of intracavitary and superficial cancers. Our purpose is to improve existing data about these sources using the Monte Carlo simulation code GEANT3. Absolute dose rate distributions in water have been calculated around these sources and are presented as conventional 2D Cartesian look-up tables. Also the AAPM Task Group 43 formalism for dose calculation has been applied. The calculated dose rate constant for the CDC.K1-K3 source is A = 1.106 +/- 0.001…
Machine Learning Regression Approaches for Colored Dissolved Organic Matter (CDOM) Retrieval with S2-MSI and S3-OLCI Simulated Data
2018
The colored dissolved organic matter (CDOM) variable is the standard measure of humic substance in waters optics. CDOM is optically characterized by its spectral absorption coefficient, a C D O M at at reference wavelength (e.g., ≈ 440 nm). Retrieval of CDOM is traditionally done using bio-optical models. As an alternative, this paper presents a comparison of five machine learning methods applied to Sentinel-2 and Sentinel-3 simulated reflectance ( R r s ) data for the retrieval of CDOM: regularized linear regression (RLR), random forest regression (RFR), kernel ridge regression (KRR), Gaussian process regression (GPR) and support vector machines (SVR). Two different datasets of radiative t…
Correcting AVHRR Long Term Data Record V3 estimated LST from orbital drift effects
2012
Abstract NOAA (National Oceanic and Atmospheric Administration) satellite series is known to suffer from what is known as the orbital drift effect. The Long Term Data Record (LTDR [Pedelty et al., 2007]), which provides AVHRR (Advanced Very High Resolution Radiometer) data from these satellites for the 80s and the 90s, is also affected by this orbital drift. To correct this effect on Land Surface Temperature (LST) time series, a novel method is presented here, which consists in adjusting retrieved LST time series on the basis of statistical information extracted from the time series themselves. This method is as simple and straightforward as possible, in order to be implemented easily for s…
Relationship between UVB and broadband solar radiation in Spain
2014
The daily values of UVB irradiation (290–315 nm), IUVB, and the broadband total irradiation (300–2800 nm), IT, measured on a horizontal plane have been correlated for the period 2000–2008 at 16 measurement sites in Spain. The results have been compared with the daily experimental values registered at the same sites during the period 2009–2011. The coefficients of determination R2 obtained by applying a linear regression are higher than 0.88 for all sites and increase to 0.94 when using a quadratic regression. When all data are considered together, the values of R2 are 0.91 and 0.97 for the linear and quadratic regressions, respectively. Three different clearness indices, which are dimension…
Formulation and test of an ice aggregation scheme for two-moment bulk microphysics schemes
2013
A simple formulation of aggregation for 2-moment bulk microphysical models is de-rived. The solution involves the evaluation of a double integral of the collection kernelweighted with the crystal size (or mass) distribution. This quantity is to be inserted intothe differential equation for the crystal number concentration which has classical form. The double integrals are evaluated numerically for log-normal size distributions overa large range of geometric mean masses. A polynomial fit of the results is given thatyields good accuracy. Various tests of the new parameterization are described: aggre-gation as stand-alone process, in a box-model, and in 2-D simulations of a cirrostratuscloud. …
Cell-average multiresolution based on local polynomial regression. Application to image processing
2014
In Harten (1996) [32] presented a general framework about multiresolution representation based on four principal operators: decimation and prediction, discretization and reconstruction. The discretization operator indicates the nature of the data. In this work the pixels of a digital image are obtained as the average of a function in some defined cells. A family of Harten cell-average multiresolution schemes based on local polynomial regression is presented. The stability is ensured by the linearity of the operators obtained and the order is calculated. Some numerical experiments are performed testing the accuracy of the prediction operators in comparison with the classical linear and nonli…