Temat: Regression model - Prolib Integro

Skocz do pozycji: 1.

Tytuł:: Потребление продовольственных товаров в зависимости от размеров домашнего хозяйства
Food consumption depending on the size of household
Autorzy:: Turczak, Anna
Wydawca:: Armenian State University of Economics
Opis:: This article presents structural parameters of linear causal-descriptive models estimated for 45 assortment groups. The computation was carried out separately for subsequent years from 2006 to 2015 and was based on data from the Central Statistical Office of Poland. The models were verified by assessing the fitting theoretical values to empirical data and checking the significance of structural parameter associated with the independent variable. As a result, food categories were determined for which linear function is a good approximation describing the stochastic relationship between the consumption per capita and the size of household as well as such food items for which the relationship is nonlinear. The first group consists of 39 out of 45 assortment items, and the second one includes only six items, namely: bread, flour, milk, potatoes, beetroots and sugar.
Dostawca treści:: Repozytorium Centrum Otwartej Nauki

Artykuł

na półce

Skocz do pozycji: 2.

Tytuł:: Sigmoid growth curves, a new approach to study the dynamics of the epicotyl emergence of oak
Autorzy:: Jastrzębowski, Szymon
Ukalska, Joanna
Wydawca:: The Committee on Forestry Sciences and Wood Technology of the Polish Academy of Sciences and the Forest Research Institute in Sekocin Stary
Cytata wydawnicza:: Ukalska, Joanna & Jastrzębowski, Szymon. (2019). Sigmoid growth curves, a new approach to study the dynamics of the epicotyl emergence of oak. Folia Forestalia Polonica, Series A – Forestry, 2019, Vol. 61 (1), 30–41. 10.2478/ffp-2019-0003.
Opis:: Three of the most frequently used sigmoidal growth curves from the Richards family are the logistic model, Gompertz model and Richards model. They are used in the analysis of organismal growth over time in many disciplines/studies and were proposed in many parameterisations. Choosing the right parameterisation is not easy. The correct parameterisation of the model should take into account such parameters that are useful to describe the analysed growth phenomenon and are biologically relevant without additional calculations. In addition, each parameter of the model only affects one shape characteristic of each growth curve, which makes it possible to determine standard errors and confidence intervals using statistical software. Growth curves in germination dynamics studies should provide information on topics such as the length of the lag in onset of germination, the maximum germination rate and, when it occurs, the time at which 50% of seeds will germinate and the final germination proportion. In this article, we present three parameterisations of the logistic, Gompertz and Richards models and indicate two parameterisations for each model, corresponding to the abovementioned issues. Our proposition is parameterisation by taking into account the maximum absolute growth rate. Parameterisations indicated as useful for germination dynamics are characterised by the fact that each parameter has the same meaning in every model, so its estimates can be compared directly amongst the models. We also discussed the goodness-of-fit measures for nonlinear models and in particular measures of nonlinear behaviour of a model’s individual parameters as well as overall measures of nonlinearity. All described models were used to study the dynamics of the epicotyl emergence of pedunculate oak. After checking the close-to-linear behaviour of the studied model parameters and by taking into account the criteria of model selection (AICc of each growth curve and the residual variance [RV]), the best model describing the dynamics of epicotyl appearance of pedunculate oak was the Richards curve.
Dostawca treści:: Repozytorium Centrum Otwartej Nauki

Artykuł

na półce

Informacja

Wyszukujesz frazę "Regression model" wg kryterium: Temat