Is neural network forecasting considered as TA? A time series projection depends on its initial condition location(chaotic is deterministic) .i.e. However, different time frames study may give different predictions. There are also both deterministic and stochastic trends in time series. In short, TA based on a single past data alone can be very subjective. We need TA that also includes other variables from the future into the equation. Multivariate(space) time series studies attempts to do so but it is not clear the cross dependence is solely relying on each others past data or perhaps a confluence of projected future results being feedback into the algo to iterate to agreement across time and space.
If you play strip poker with these pro t/a merchants , you will uncover their true agenda , most of them are probably marketing/selling something ,but you will never get any proof with t/a studies.
For me TA is as follows: Analysis of the markets using just Price and Volume with absolutely zero regard for news and fundamentals, relying on repetitive patterns at areas of interest.
That was my problem today. I took 1/2 my profit from my TA setup instead of letting price reach my target since I wanted to be correct more than rich today. Over the long term this is terrible risk vs reward. I just need to keep reminding myself that. Quote from jack hershey: It will always be true that most people would rather be correct than rich. . [/B]