ICA PREPROCESSING FOR TIME SERIES PREDICTION
Simona M l roiu, Kimmo Kiviluoto and Erkki Oja, Helsinki University of Technology,
Simona.Malaroiu@hut.fi, Kimmo.Kiviluoto@hut.fi, Erkki.Oja@hut.fi
We propose a new method to predict time series us-
ing the technique of Independent Component Analysis
(ICA) as a preprocessing tool. If certain assumptions
hold, we show that ICA can be used to transform a set
of time series into another set that is easier to predict.
These assumptions are not unrealistic for many real-
world time series, including nancial time series. After
prediction, the original time series can be constructed
by the inverse ICA transform. We have tested this
approach on two sets of data: arti cial toy data and
nancial time series. The results suggest that these
can be predicted more accurately using the ICA pre-
processing.