Mathematics in Education, Research and Applications (MERAA), 2019(5), 1
Back to Content
-
Published online 2019-10-15
DOI:https://doi.org/10.15414/meraa.2019.05.01.1-8
Nonparametric distribution of the daylight factor
Dušan Páleš, Milada Balková
Slovak University of Agriculture in Nitra, Slovak Republic
Article Fulltext (PDF), pp. 1–8
- Kernel density estimation (KDE) approximates the distribution of statistical data similar to the histogram. The histogram of data
is a special kind of the Kernel density. In the reconstructed building of stall in Oponice (Slovakia), we measured the values of
daylight factor. The obtained data proved a bimodal distribution, so it was not appropriate to use some of the usual parametric
distributions. This paper describes how Kernel density can be applied to measured results. We find out the values of the
cumulative distribution function of such density, by probability procedures, that serves us comparison with the prescribed values
of the daylight factor in the standard, on the one hand for animals (1.0%) and on the other hand for the people (1.5%) who care
for animals. The results obtained from the measurements and the same ones approximated by KDE are in good agreement.
- Keywords: Kernel density estimation, bandwidth, daylight factor
- JEL Classification: C13, C16