Skip to Content

Estimating the embedding dimension distribution of time series with SOMOS

Congresses name: 

International Work-conference on Artificial and Natural Neural Networks, IWANN 2009


Salamanca, Spain

September, 2009

The paper proposes a new method to estimate the distribution of the embedding dimension associated with a time series, using the Self Organizing Map decision taken in Output Space (SOMOS) dimensionality reduction neural network. It is shown that SOMOS, besides estimating the embedding dimension, it also provides an approximation of the overall distribution of such dimension for the set where the time series evolves. Such estimation can be employed to select a proper window size in different predictor schemes; also, it can provide a measure of the future predictability at a given instant of time. The results are illustrated via the analysis of time series generated from both chaotic