probability of up in selected stocks by both models 0.6*0.7=0.42
probability of up in selected stocks by model but not anoth model 0.6*0.3 +0.4*0.7=0.46. the stocks selected may up or down
probability of down stocks 0.4*0.3=0.12.
The gain here is that stocks to go down is decreased.
the key point here is whether the two models are correlated or not? I believe most indicators are not independent, but somehow correlated as they all are derived from the same price action, so it's not proper to calculate the combined winning probability by treating them as independent model or indicators.
the key point here is whether the two models are correlated or not? I believe most indicators are not independent, but somehow correlated as they all are derived from the same price action, so it's n ...
Seawave 发表于 2009-12-13 11:56