4 ms·
Do not minmax scale the data. New data outside ranges seen before will be chopped off. Do make relative (high/low), so patterns you find generalize to other dom
by zugweg 8y ago
Do not minmax scale the data. New data outside ranges seen before will be chopped off. Do make relative (high/low), so patterns you find generalize to other domains. Don't fit complex LSTM models on tiny datasets. Somehow this is the shiny thing all starters jump on, while the most simple models are the most proven and robust. Don't just share such a model without giving a disclaimer ("I never tried this out on real data, and would not invest my own money in this solution"). This is real money, real economies, and you could horribly crash these. That is also why you don't write with authority about how to build a bridge (while you never crossed a bridge you've build before). The only reason for the "clearly showing how powerful LSTM models are" is it responding to test set samples it got wrong, and correcting with a lag. How about some evaluation, preferably on another stock than trained on?
- rl3 8y ago>This is real money, real economies, and you could horribly crash these. That is also why you don't write with authority about how to build a bridge ... Setting aside that first bit of hyperbole, I don't see the danger in the bridge example. I mean, if engineers are building bridges based on information they've gleaned from internet articles, there are larger issues at play. As far as markets, retail traders basically exist and have existed within a massive bubble of misinformation since forever. There's a reason order flow is so valuable.
- zugweg 8y agoML is dangerously overhyped (that is an issue at play). This article uses min max scaling, so it will treat a 30$ stock as if it were 20$. You want someone like that managing your pension fund? I refer to engineering code, because it is clear to me we don't want our bridges build by those not skilled enough to make safe bridges. Yet, it is not so clear with ML/AI, while the potential for damage may be even greater. https://en.m.wikipedia.org/wiki/Financial_Modelers%27_Manifesto https://en.m.wikipedia.org/wiki/Financial_Modelers%27_Manife...