On this forum, only seven answers about neural network research !
did someone use a neural network with ddl in multicharts ?
Thanks
neural network
-
sptrader
- Posts: 742
- Joined: 09 Apr 2010
- Location: Texas
- Has thanked: 483 times
- Been thanked: 274 times
- Contact:
Re: neural network
While on the subject of Neural Networks, I'd love to see MC take the lead on NN in technical analysis software.
It would be a new market for them. With all of the AI rage these days, it's time someone offered it.
Make it an optional plug-in module that we could buy.
The product that I used to use was by Ward Systems Group, the software was called Neuroshell2.
I gave up using it when everything went 64bit. It was a 32 bit program that I could call from TS or MC using a DLL.
Steve
It would be a new market for them. With all of the AI rage these days, it's time someone offered it.
Make it an optional plug-in module that we could buy.
The product that I used to use was by Ward Systems Group, the software was called Neuroshell2.
I gave up using it when everything went 64bit. It was a 32 bit program that I could call from TS or MC using a DLL.
Steve
-
Smoky
- Posts: 518
- Joined: 03 Dec 2010
- Location: Thailand
- Has thanked: 99 times
- Been thanked: 121 times
Re: neural network
i only find a c++ library "FANN" who can be used with MC.While on the subject of Neural Networks, I'd love to see MC take the lead on NN in technical analysis software.
It would be a new market for them. With all of the AI rage these days, it's time someone offered it.
Make it an optional plug-in module that we could buy.
The product that I used to use was by Ward Systems Group, the software was called Neuroshell2.
I gave up using it when everything went 64bit. It was a 32 bit program that I could call from TS or MC using a DLL.
Steve
-
sptrader
- Posts: 742
- Joined: 09 Apr 2010
- Location: Texas
- Has thanked: 483 times
- Been thanked: 274 times
- Contact:
Re: neural network
Smokey: Here is a .dll example that I used, to get the NN output into MC.
DefineDLLFunc:"TRADES.DLL", FLOAT, "TSGetOutput", INT;
DefineDLLFunc:"TRADES.DLL", FLOAT, "TSGetOutput", INT;
-
Smoky
- Posts: 518
- Joined: 03 Dec 2010
- Location: Thailand
- Has thanked: 99 times
- Been thanked: 121 times
Re: neural network
Nice, what tools do you use to make your NN ?
You can build, train, test a NN with this freeware
https://firagiel.com/web/technical-soft ... l-network/
and then use FANN library code to load your NN and make prediction.
You can build, train, test a NN with this freeware
https://firagiel.com/web/technical-soft ... l-network/
and then use FANN library code to load your NN and make prediction.
-
sptrader
- Posts: 742
- Joined: 09 Apr 2010
- Location: Texas
- Has thanked: 483 times
- Been thanked: 274 times
- Contact:
Re: neural network
Smokey : NNets are very difficult to get working the way we want. You need a target and inputs to predict that target. There are infinite possibilities of each. One important thing that I learned, is that you Don't predict price directly, you predict something that is correlated with the price target.
Its been 15+ years since I went down that road, so I don't remember too much detail about it. You might try a google search for NNet inputs and targets, for the price that you're trying to predict. It was a fun project but I never came up with a trading system that was reliable enough, to trade with. With today's AI however, it might be more possible today.
Its been 15+ years since I went down that road, so I don't remember too much detail about it. You might try a google search for NNet inputs and targets, for the price that you're trying to predict. It was a fun project but I never came up with a trading system that was reliable enough, to trade with. With today's AI however, it might be more possible today.