Machine learning in trading: theory, models, practice and algo-trading - page 173
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
All so, but this pattern is very bearded, everyone knows about it for a long time, so...
And now the simplest classifier would have found it, for one series such interpretation of price, volume and OM, not enough data, you need at least orderbook and strip with directions of deals, and in the case of forex, because this information does not exist, you need to take it from the Western liquid futures markets.
With real volumes the situation is similar.
Unfortunately I do not have the right and frankly wish to go deeper, after all we take money from each other)))) But 2 years ago, when I was working with SAM, I was processing more than 500 chips at the input of the neural network and had about 30 outputs, but time goes... ;)
The regularity is bearded, but it is valuable because it has a "physical meaning". The growing OM shows that money goes into this price movement! And they are the fuel that moves the price. OM is not a derivative from a chart (like technical indicators, for example), it is an independent parameter. It is impossible to cheat. If there is an inflow of money - the OM goes up. Money is flowing out of the market - OM starts falling. Everything is clear and interconnected, practically a market law. Therefore OI can and should be used. But not everyone knows how. And in Forex, it is not even directly available, as you have correctly written.
With real volumes, the situation is similar.
The regularity, though bearded, but it is valuable because it has a "physical meaning". The growing OM shows that money goes into this price movement! And they are the fuel that moves the price. OM is not a derivative from a chart (like technical indicators, for example), it is an independent parameter. It is impossible to cheat. If there is an inflow of money - the OM goes up. Money is flowing out of the market - OM starts falling. Everything is clear and interconnected, practically a market law. Therefore OI can and should be used. But not everyone knows how. And in Forex, it is not even directly available, as you have correctly written.
With real volumes the situation is similar.
Come on, I take the volume and OI from the pound futures with CME, I also take the real volume in real time from the delta cluster. So everything is available, and the difference between futures and spot is only a few pips, so.....
Is it more profitable to trade futures than to trade on the spot?
Who would argue? The OM is a very important feature, one of the most important, just the patterns have become a little "trickier", not so directly in the forehead as stated above by respectedMihail Marchukajtes
So the chart there says, EXPECT a reversal, for example today on the pound. Volume is down, OI is down, exchange rate is down, recommendations, gain strength to the upside. This does not mean that today we will go up, it means that we have to look for an entrance point to buy.) I bought a little bit from the bottom, and before that I sold because the TS, you can not go against it.... Well, again, although there is a mistake, circled in red, but still overall very good, and the purchase from below in accordance with the recommendations on the OM, I think even hold it longer, if the stop will not knock it out, because on Monday I think the pound will continue to grow, and very significantly...... IMHO
I think there is not much difference if the pairs are moving in unison.
The analogy about 3 pixels and HD is very relevant in my opinion to what most of us are doing here.
Profitable models are obtained not even on 10000 and not on 500 predictors, but less, say, on 10. It's not a big deal to put 500 predictors into NS. Noises will drown everything out there. I'm not surprised that it didn't work well.
90% of the problem (the other 10% is to form predictor candidates and choose the MO method) is unbiased model selection. Somehow the bottleneck is often in how one interprets the training results and applies them.
I'm sure that even with lots and lots of information, screwing up an experiment can be very easy. And it's not about the method or the data (which will be noisy anyway), but about how to separate a model trained on noise from a model trained on signal.