Machine learning in trading: theory, models, practice and algo-trading - page 2442
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
Maybe someone will be bored.
https://www.econometrics-with-r.org/16-4-volatility-clustering-and-autoregressive-conditional-heteroskedasticity.html
First, wait for the beta versions of matrix operations
What is the approximate timeframe?
Are there any plans to load models saved in tensorflow?
We have already said that we are moving towards implementing machine learning in MQL5.
Soon we will release native support for complex numbers (ready), speed vectors and matrices. This is a native functionality of the language, not a library.
Next, we will include a large set of ML mechanics and give functionality similar to TensorFlow. This will allow you to write a completely different level of native robots.
Thatsounds really cool, if they make a similar scale fit with data it would be mega-cool, it would reallysurpass all sorts of Robinguds and Ninge. Otherwise you can't do much good with just plain banged up, non-synchronous candlesticks\there's not much else to do. What I mean is:
Just write the whole market second by second, synchronously with a couple hundred (or at least dozens) of different real economic indicators, and write the news with a timesample, too. That would be the bomb! All this practices from small to large write themselves, but makeshift, who how, somefragments of the general information background, and you could make anextra class marketdata-server, even sell quality history, if at reasonable prices, since everyone has long been clear that the whole alpha in the data, and not the newfangled algorithms ML, "garbage in the input garbage out" and ip.
Good luck!
The work in this direction is in progress, we are preparing an independent data storage on tens of thousands of macroeconomic indicators.
It will be available on all MT5 terminals regardless of brokerage accounts in addition to brokerage flows.
The work in this direction is in progress, we are preparing an independent data storage on tens of thousands of macroeconomic indicators.
It will be available on all MT5 terminals regardless of brokerage accounts in addition to brokerage flows.
This is interesting. Will you restore the history, or will you be collecting data from a certain point? Will there be an automatic synchronization with the current chart, taking into account the different moments of time in different countries in different years?
That's interesting. Will you restore the history, or will you collect data from a certain point? Will there be an automatic synchronization with the current chart, taking into account the different moments of time translation in different countries in different years?
With full history and in UTC time.
Most likely, it will be possible to set the time in the terminal settings for automatic adjustment.
When will the meta-editor itself be improved? Why not take pycharm as an example and just copy its functionality?
The work in this direction is in progress, we are preparing an independent data storage on tens of thousands of macroeconomic indicators.
It will be available on all MT5 terminals regardless of brokerage accounts in addition to brokerage flows.
The work in this direction is underway, we are preparing an independent data storage on tens of thousands of macroeconomic indicators.
It will be available on all MT5 terminals regardless of brokerage accounts, in addition to brokerage flows.
Respect!
Sounds like a fantasy, I didn't even think it would ever appear. In fact for machine learning the most important thing is to have access to different indicators with strict periodicity of publications. Reliable storage of this data is the solution to the problem in my opinion!
Yes, data is almost everything in our difficult business, in serious organizations most of the data is gathered independently, it's a separate large-scale layer of work, single algorithmic traders and mini teams cope with this task mediocrely, which is mostly the cause of frustration in algorithotrading, and "ordinary traders" don't go further than perversions with one BP, which is a baby steps with predictable results. If we had a professional organized data logging for MT, it would be a huge competitive advantage, and it would give clients a chance to try themselves on the real battlefield, but not in a child's sandbox.