Machine learning in trading: theory, models, practice and algo-trading - page 2358
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
Go to the Alpari calendar, click on the news about the release of the current indicator values - More information about... - a table of historical data on this indicator with publication dates
For what period of time is the information available there?
Lots to download/spare thanks to the old Fed)
https://www.federalreserve.gov/data.htm
somewhere there was even an opportunity to immediately compile a time-synchronized table of heterogeneous quantities by hand and download
What period of time is the information available there?
You have to look at the depth for each indicator.
For example, Schweyaria's unemployment is up to mid-2007 in depth/.
Yes, you can download
You have to look at the depth of each indicator.
For example, Schweyaria's unemployment is up to mid-2007 in depth/.
Yes, you can download
Normas, is there only access through the LC?
Ah, well found it, they have a calendar from fxstreet
Lately we have been studying the works of Prado.) Therefore, I will reply with a quote from his book: "The fundamental data are extremely regularized and low-frequency. Given the public availability in the marketplace, it is unlikely that there is any value in them for exploitation."
It is fundamental data that is causing all the trends. Over the past few years I can recall strong trends in commodity currencies (and commodities) and the Eurobucks. And commodity currencies plowed through it quite classically.
Any simplest-ancient strategy on any goddamn oscillator can squeeze the maximum profit out of it for 5 years ahead.
But as already mentioned, there is always something new mixed in with the fundamentals and therefore the classical models of classification/regression by history are not useful, which is probably what Prado meant), (moreover there are few big trends) And most likely they degenerate to 2-3 most significant indicators. Thereal task comes down to processing the entire news background and it is probably even closer to AGI.
All trends are caused by fundamental data. Over the past few years I remember powerful trends in commodity currencies (and commodities) and the Eurobucks. And the commodity currencies plowed through it quite classically.
Any simplest-ancient strategy on any goddamn oscillator can squeeze the maximum profit out of it for 5 years ahead.
But as already noticed, there is always something new mixed in with the fundamentals and therefore the classical models of classification/regression by history are of little use, which is probably what Prado meant), (the more so there are few big trends) And most likely they degenerate to 2-3 most significant indicators.The real task comes down to processing the entire news background and is even closer to AGI probably.
To use FA, you have to spend a lot of time to digitize systematically for machine processing. This is feasible for the collective. For an individual it would take years. I know how to do this, but one cannot do it alone.
Can MO calculate signal quality by probability? To then filter out those with more than 90% probability
Can MO calculate the quality of signals by probability? To then filter out those with more than 90% probability
Maybe. Trees and forests. You can select leaves with only >= 90% class on the training plot. But on new data, they will give a 50% error. This is for teacher markup with TP=SL. Other types of markup can also be sampled at the right percentage. But profitability is likely to be zero, as in my case.