"ChatGpT like" machine learning AI

acrigney
28 Apr 2023, 00:59

RE:

Dear Alex,

                   I have developed a generic ML framework using Microsoft's ML.NET that uses templated types called generics. Not sure of your programming experience but that means that if you give me a data stream I can very quickly build models. With Python ML there is no generic interface for the ML libraries so they have to copy and paste their code for every problem. This makes it very hard to build generic solutions! I have used my ML framework for trading for about 2 years now but you also need the catching/recovery/filtering options which are just as important. However I found that the minute data was not accurate enough with ctrader and I have tried using the tick data but this gave models that converged nicely but gave inconsistent results. So now I am back to using minute data but I have added renko data and also derivatives. I have many types of models too, but I have not used models on multi timeframes and maybe I should do that. There was another problem that it was a huge change to move to .net 6. I have had to rebuild a very large amount of code, it should have been easy but many issues had to be fixed and also I am still unable to attach the debugger to my bots and I have complained to ctrader many times but they just say it works and a lot of other people have this problem too. I don't think chatgpt is going to work unless you are thinking of long term trades as in over days or hrs. I think the longest timeframe you can trade is a minute as the market can move so fast. But I want to incorporate chatgpt too as well maybe for stock/options trading. I trade on the index markets.

I would love you to join me in my efforts and testing, its been very hard to find traders who are interested in using ML!! My ML uses AutoML, that means that it will find the best mode and auto feature engineering while the model is being built. Also it can tell you why it has made the prediction for every prediction. 

Best Regards,

                         Alistair

alexsanramon said:

Good Day,

 

I hope this message finds you well. I am writing to respectfully request your assistance on building a simple indicator that has machine learning capability. Suppose a momentum indicator that has been fed with 2 years of machine learning data and can forecast only 1 bar of future data.

I understand that your time is valuable, but I would greatly appreciate any guidance or insight you can provide on this matter.

Thank you for your consideration.

Sincerely, 

Alex

gnsmaciej
23 Dec 2024, 09:42

Is it possible to teach the bot, for example, through a very large number of profitable transactions presented visually, where there are several indicators on the charts, or on this basis you can program the bot to draw conclusions / assumptions of entry and exit by itself ?

acrigney
23 Dec 2024, 10:08

RE: "ChatGpT like" machine learning AI

Guys this idea is really like looking at tea leaves, you can see anything in a historical chart as the human tries to make sense of the randomness of it. But really the ML should be able to take into consideration far more numbers of features that a human can (such as price action) and hence overall in theory you would expect it to win. But you have to be trading over short time frames (relative to your market) so that external factors or manipulation of markets have less impact. My ML has around 800 features to build models from but I am refactoring the features into a reduced set to reduce overfitting of the models. Also you need to support very good trading adaptability to dynamically adapt trading strategies to suit market conditions and it goes without saying that whatever algorithm you use you must have excellent equity management, appropriate types of orders, hedging strategies etc. All my bots, ML/grid have this as standard.

gnsmaciej said: 

Is it possible to teach the bot, for example, through a very large number of profitable transactions presented visually, where there are several indicators on the charts, or on this basis you can program the bot to draw conclusions / assumptions of entry and exit by itself ?