The paper reviews the statistical methods of time series analysis used in a selection of papers from respected scientific journals. In particular, problems are considered in the search for cycles, the use of regression to establish causal links between variables, transfer function modelling and the use of filtering to extract componentes of time series.
An attempt is made to assess how useful the ideas of ARMA and Transfer Function modelling might be in improving the efficiency of statistical inference in these contexts.
@article{urn:eudml:doc:40029,
title = {Problems in scientific time series analysis.},
journal = {Q\"uestii\'o},
volume = {8},
year = {1984},
pages = {9-19},
language = {en},
url = {http://dml.mathdoc.fr/item/urn:eudml:doc:40029}
}
Tunnicliffe Wilson, Granville. Problems in scientific time series analysis.. Qüestiió, Tome 8 (1984) pp. 9-19. http://gdmltest.u-ga.fr/item/urn:eudml:doc:40029/