I have a data set of biological signals (PSG signals); the dimension of the signals is high (850 features for each sample). I am looking for the best way to reduce the dimensionality of the signals. I tried to down-sample:
I have also looked through PCA (Principal Component Analysis), but is it a good when applied to raw data of the signals?
What are the best algorithms used for dimensionality reduction and compression of time series data represented by PSG signals (such as flow signals, thoracic signals)?