Forgot... I try to predict at t+1 with AR model and Trigonometric model too. Seems prediction are ok too. But i don't ty to make some stats to know how good they predict.
Did someone try to compare ? : - DT close cross - linear predictor - AR predictor - TrigFit predictor (should be very powerfull for StochD at t-1 for example !) - Non linear/non continuous predictor : i mean by this a predictor based on threshold taking in consideration : angle between the two curves, space between the two curves, etc... We can imagine that for a large angle and a small space, there is a good chance the cross will occurs - some neurals net maybe ??? or another technics from the SF world but which will be common tomorrow ; ) ? I expose just ideas to share. Cheers, Mich. --- In [email protected], "john_dxd_smith" <[EMAIL PROTECTED]> wrote: > > > Dimitris Tsokakis is live and well. > > See his article "Anticipating Moving Average Crossovers" in February > 2007 issue of Technical Analysis of STOCKS & COMMODITIES. > http://www.traders.com/Documentation/FEEDbk_docs/Abstracts_new/Tsokaki s/tsokakis.html >
