Barhopper,
Sorry I was not trying to say that you should move to usual technical analysis, I was meaning that as you can't use this method directly to trading as curve, you need to find some specific ways of using this kind of techniques to trading.
Because the signal processing algos are eliminating noise and therefore making the past moves more simple, one real world trading implementation possibility might be to apply some shape recognition/statistical algorithms to the generated curves to approximate next direction or length.
The problem with statistical methods is that the amount of data would need to be quite large and that would probably restrict you to use lower timeframes that are (IMHO) quite unpredictable/unsuitable for trend style trading.
ps. to be honest, I'm not quite sure how this is so different from moving averages, the lag with moving averages is not really as critical as is the fact that this method repaints the whole curve at every point when new data arrives.
Sorry I was not trying to say that you should move to usual technical analysis, I was meaning that as you can't use this method directly to trading as curve, you need to find some specific ways of using this kind of techniques to trading.
Because the signal processing algos are eliminating noise and therefore making the past moves more simple, one real world trading implementation possibility might be to apply some shape recognition/statistical algorithms to the generated curves to approximate next direction or length.
The problem with statistical methods is that the amount of data would need to be quite large and that would probably restrict you to use lower timeframes that are (IMHO) quite unpredictable/unsuitable for trend style trading.
ps. to be honest, I'm not quite sure how this is so different from moving averages, the lag with moving averages is not really as critical as is the fact that this method repaints the whole curve at every point when new data arrives.