Kreslik.com - traders community since 2006

GA primer (strongly recommended):
http://www.cna.org/isaac/gaprim.pdf
or locally: GA primer

Bibliography on Genetic Programming:
http://liinwww.ira.uka.de/bibliography/ ... mming.html

Intro to Genetic Algorithms (simulated annealing mentioned, too):
http://www.burns-stat.com/pages/Tutor/genetic.html

TS research genetic tools for Tradestation:
http://www.tsresearch.com/download/free ... izer_free/

Optimax genetic tools for Tradestation:
http://www.modus-novus.com/optimax/

last updated: 7/2/2006

Please feel free to suggest additional important links relating to this subject.

Hi Michal,

this link "GA primer (strongly recommended):" is not available.

Thanks

DavidF

Hi, David,
I have noticed no issues downloading & viewing the pdf file from the location specified. However, I uploaded and attached the file over here and added the "or locally" link, too. Please PM me if you got further troubles downloading or viewing the locally stored file. Tnx

Hi Michal,

I have read important link in this section about GA. But I do not sure that I good understood how best way I can use GA for my trading. Maybe other trades like I am have some problem and any questions about it. Can you giv me simple description or small example of GA and trading.
Tnx DavidF

Hello, David,

genetic algorithms are not used in trading directly. They help us to arrive at the best model (model parameters) and variable values (value parameters).

Model parameters determine what concept or model will be used in the strategy in question. Example:

- we are not sure which moving average calculation will be the best one to use with our trading strategy, so we design an input like iMAType. Let's say that the range for this input parameter will be 1 to 4. According to the selected input value (1, 2, 3 or 4) we calculate the moving average within the strategy either as a simple, exponential, weighted or triangular one.

Value parameters are pretty self-explanatory. Their value is used as an input to already existing model. Example:

- we want to know what moving average length is the best for our strategy, so we design a variable like iMALength. This variable is used in the strategy code directly as an input for calculating the moving average length.

The two main advantages of GAs are:
- their fitness landscape is shaped gradually, because they don't try all possible combinations. This means that GAs are not easily fooled into local maxima
- the solving speed is tremendous compared to the standard exhaustive methods

I use GAs in my system development every day. My best strategies are designed using GAs. Of course, GAs are no holy grail - the system developer must still know very well what is he doing and put a lot of hard work into the development.

The markets have become more efficient since 1990ies due to the widespread computer analysis software. If you want to gain an edge, it's necessary to use computer aided system development tools like, for example, GA.

Another interesting AI tool used in trading strategy development might include back-propagated neural networks, but that's another topic

Hi Michal,

Great links. The bad part is I now realize how little I know about the phenomenal advancements made in the last few years in regards to retail trading solutions.

As far as the current Global Optimization software solutions available, I was hoping to find something to piddle around with while I read. However, after a bit of research I am finding the realistic solutions (at least the one's that have not been sold to large hedge funds) have a steep barrier to entry in both a financial and learning curve sense. No doubt this will be eventually solved via further integration into the mainstream populace. I am hoping that by following some of the forum member's advancements with NeoTicker and/or MT4 I will be able to tag along with it's evolution.

A couple of standalone versions I have been looking at are:

www.tradingsolutions.com of which I cannot find much feedback on and www.neuroshell.com that can be autotraded via Interactive Brokers. I saw where Whitmark had been using NeuroShell Trader to compare Neoticker Grid Optimizer output. However some users have reported it's lack of graphic tools to be a huge detraction. Both priced at over $2k for a real time platform. Trading Solutions at least has a 30-day free trial I am tempted to mess around with.

Another solution that seems to have a decent following is www.tradingblox.com but it does not yet have a real time data feed option, thus limiting it to longer term EOD trading.

Yet another is www.tradingsolutions.com which is a 2006 Reader's Choice in Stocks & Commodities Magazine.

Not sure if I am telling you anything you don't already know here, but I suppose at this point the best tactic would be to watch the two frontrunning charting packages as determined by forum members, MT4 and Neoticker, to see which can implement a satisfactory GA solution.

My problem is I have this beautiful trading system I have explored graphically with a few indicators that I am just chomping at the bit to optimize. LOL.

Forest

Forest,

you are talking about the neural networks solutions here.

Neural nets are a subset of AI, as well as genetics is. Personally I am not wild about neural nets since

• they are prone to curve-fitting
• you can't see the underlying logic of neural net-generated solution

Michal

Forest,

I don't know too much about Neural Nets myself since I primarily utilize GA, however, I strongly recommend you look into Mark Jurik's Braincell, which is a neural net utilizing back percolation (instead of the traditional backpropogation which can be prone to curve fitting), and DDR which helps you select uncorrelated indicators and cleans up your data for your Neural Net. Although I haven't used Braincell myself I have heard of others using it with success and I have also used Mark's other products and enjoy them immensely.

Mark's website http://www.jurikres.com/ has a great deal of information regarding noise reducing indicators, and he himself advocates using neural nets to train indicators (as opposed to the normal methodology of using neural nets to predict price movement).

Additionally, if you are looking for more academic research regarding neural nets used in the markets check out http://www.smartquant.com/references.php. Dr. Anton Fonton who runs Smart Quant has a great reference library for neural nets and other interesting things for finance. He has told me that some of the links are not currently working because their site was hacked but he says that they are working on bringing things back online.

Edward

Michal,

I hope this doesn't drive things too off topic, but do you have any links for resources regarding Simulated Annealing? The ones at SmartQuant are offline.

Edward

Hello Edward,

Thanks for the wonderful suggestions. Quite a few years ago I purchased Jurik's indicators for TradeStation. I didn't realize he had developed a Neural Net software application since I have been somewhat out of the loop for a couple of years.

I will also browse Dr. Fonton's site and appreciate the reference.

Forest