Ambient error weighted partial least-squares regression: a new receptor model
✍ Scribed by Shizoom Wang; Timothy Larson
- Publisher
- Elsevier Science
- Year
- 1993
- Tongue
- English
- Weight
- 464 KB
- Volume
- 272
- Category
- Article
- ISSN
- 0003-2670
No coin nor oath required. For personal study only.
✦ Synopsis
In the proposed model, both the amblent data and the source profiles were weqbted by the amblent mass scaled amblent error The comparison between thts model and the effectwe varrance we&ted chemtcal mass balance-type model was done usmg NBS test data sets The compared results show that the amblent error wezghted par&l least-squares regresston model was better tn predlctmg the emisszons from each source However, the effectwe vanance welghted chemical mass balance model dzd better m predlctmg the amblent ~ncen~ation of each species, which can easily be understood from the formulation of these different approaches Ktqwords Partial least-squares regresston, Receptor models