3 ms·
You're presenting this from the perspective of the author's selection of MA(3), but I think what you're really getting at is yes, MA(3) happened to be comparabl
by longerthoughts 8y ago
You're presenting this from the perspective of the author's selection of MA(3), but I think what you're really getting at is yes, MA(3) happened to be comparable to the ML model in this case, but the beauty of the ML model was that it knew to nearly replicate MA(3) out of a great many common indicators. Is that essentially what you're saying? If so, I'm curious how things like longer period moving averages or an exponential moving average perform relative to both MA(3) and the ML model in this case.