Figure 3 - uploaded by Boris Bonn
Content may be subject to copyright.
Download
View publication
Proportion of events and errors in a sliding window of width 60 days.

Proportion of events and errors in a sliding window of width 60 days.

Source publication
Figure 2: Nucleation parameter and proportion of events along the first...
Figure 3: Proportion of events and errors in a sliding window of width...
Figure 4: Regression weights in a sliding window of width 100 days. The...
Figure 5: Proportion of events and errors in a sliding window of width...
Data mining approaches to explaining aerosol formation
Article
Full-text available
  • Jan 2006
Atmospheric aerosol particle formation is frequently observed in various environments. Yet, despite numerous studies, processes behind these so called nucleation events remain unclear. In this work we describe the use of data mining techniques to detect factors influencing particle formation. These techniques are applied to a dataset of eight years...
Cite
Download full-text

Context in source publication

Context 1
... classification together with feature selec- tion on the whole data set has given us the two variables, relative humidity and the condensation sink, capable of explaining 88% of the nucleation events. However, seasonal variation of all variables is strong, and nucleation also has a strong seasonal behaviour (see Figure 3). Hence, it is natural to ask whether explaining the events would be easier if we only wanted to do it in one season. ...