Dakota

[slogem]

Does anyone here have any experience with the Dakota software from BioComp? I have been reviewing the website and it looks interesting: no indicators, rather a swarm of neural trading systems (they call them bots) that pursue their own individual trading pattern. The most productive of the bots control the current trade. (At least this is how I understand the process.) The system is claimed to be totally unoptimized and walk-forward which is an amazing claim. The entry cost is not huge, about $590, but there is no trial period or price. So the trial is $590. They are pretty much into take-it-or-leave-it. All they offer are testimonials which of course are predictable in their praise. I wouldn't mind making the investment if I had contact with someone using the system. One other aspect that puts me on edge is that the emphsis is on stocks, and I am into trading futures.

 

[Pat494]

Too expensive for me to find out if they are any good. So then one reads the advertising blurb.
I find that what is not said is sometimes even more relevent than what is said. After all they want the product to appear in the best light without actually telling lies.
Are the figures they give any good ? Which figures do they give and which do they conveniently miss out ?
Hope you will post your observations if you buy it.

 

[slogem]

Hope you will post your observations if you buy it.

I will certainly do so if I ever get that far. I have the message on several boards, but no Dakota user has popped up yet.

 

[Pat494]

I have just had a look at their website. They do actually give figures for S&P 500, which looks like a steady income (equity ).

 

[Pat494]

I said I would let you know how I got on.
I couldn't resist buying it ( I'm a sucker for complex systems and pretty women ).
Had it for a week.
Cost US$549 (but with a15% discount.). There is a cheaper version but it doesn't have the build your own Bot bit.
I concentrate on the Dow Jones 30 and have the OHLC + volume ( dodgy ) figures to play with.
There is only one GOOD Bot that comes with it. The results from the others on the Dow are mostly negative !!!
So from March 10 2000 it did pretty well to notch up 11,474.7 points. That is even more impressive because it is not curve fitting etc.
I have struggled through a Visual Basic tutorial to write 2 more Bots. I even have one or two ideas that might make good Bots, but not able to construct them yet.
Well that's it so far. Let me know if this is of some interest.
:?: :?:

 

[slogem]

I said I would let you know how I got on.
I couldn't resist buying it ( I'm a sucker for complex systems and pretty women ).

Glad to hear about your experience. Looks encouraging. I got the announcement for the Express version and I am leaning toward giving it a try at the $300 price. I can then upgrade to the full blown version if the performance is adequate. Sounds like there is reason to believe it is. Response from a real user was what I was looking for.

Did you find any reference for futures...metals, grains, interest rates, currencies, softs? What data source are you using? Does Dakota use ASCII or Metastock format for the input data?

Thanks for the note

slogem

 

[Pat494]

Glad to hear about your experience. Looks encouraging. I got the announcement for the Express version and I am leaning toward giving it a try at the $300 price. I can then upgrade to the full blown version if the performance is adequate. Sounds like there is reason to believe it is. Response from a real user was what I was looking for.

Did you find any reference for futures...metals, grains, interest rates, currencies, softs? What data source are you using? Does Dakota use ASCII or Metastock format for the input data?

Thanks for the note

slogem

I have been a Metastock user now for some years and use their filing data system which works well with Biocomp. I used to use just one csv file in Excel to hold all my data and used that some years ago. It worked but seemed to confuse Biocomp using both. I have put the data into Ms manually to save the extra expense of a data provider. Overall it is ok but I had to occasionally estimate highs and lows and volume.
There is a free knowledge base where people have used Dakota on a wide variety of commodities and stocks, but I am only really interested in the Dow and S&P. There is a tutorial on making one's own bots with which I am presently struggling.
Best of luck

 

[Kunal]

Pat, there are some excellent books on VB which will put you on the right track. "VB for Dummies" is very basic and probably more than enough for what you need. The name is very misleading.
Another good one is by Deitel.

This is from the BioComp site.
There are NO technical indicators needed within Dakota. Just use the price series of the security traded.
Dakota does not regress or "back test", it adapts its trading bots as they move forward through time.
Which means what ???
If it uses OHLC relationships, curve fitting IS involved, no matter what they say.

 

[Pat494]

Pat, there are some excellent books on VB which will put you on the right track. "VB for Dummies" is very basic and probably more than enough for what you need. The name is very misleading.
Another good one is by Deitel.

This is from the BioComp site.
There are NO technical indicators needed within Dakota. Just use the price series of the security traded.
Dakota does not regress or "back test", it adapts its trading bots as they move forward through time.
Which means what ???
If it uses OHLC relationships, curve fitting IS involved, no matter what they say.

Yes I may have to get the VB book. It certainly isn't easy to understand, I look forward to many hours puzzling over it. The Bots seem to change as they walk thru the data one bar at a time. Just using the close of the Dow however is hopeless. Early days though.

 

[Pat494]

As an update which may be of interest. I had an idea of having a long ma - short ma and comparing that with yesterday's long ma - short ma. With some help from Carl Cook at Biocomp I surprised myself in being able to make a new bot !! ( but not easy).
On the S&P500 OHLCV it made 400+ points over a 6 year period. Most of those points were made in the last year or two, so pretty good I think. As I put the daily figures in I can run saved systems along 1 bar to get the latest forecast.
(I am only a customer of Biocomp and not trying to sell it etc.)

 

[mxtrader]

Never heard of this program before, so how is its performance after 2 and half year after you bought it?

 

[Pat494]

Never heard of this program before, so how is its performance after 2 and half year after you bought it?

If you can program in VB you may have a chance. For me personally it sits on the shelf now

 

[jalsck]

I am a long time user of Dakota and have put together a high level description of how the program works. The key strengths of the application are discussed. A form of optimization does take place via the swarm of trading robots, however, all signals are out-of-sample. This assumes that the parameter ranges have not been optimized over the trading simulation period.

BioComp Dakota is a product of BioComp Systems Inc. BioComp Systems provide state of the art modeling, prediction and optimization technologies to corporations and individuals. Dakota is a stand-alone application for building trading systems. The Dakota application framework is flexible thereby, enabling trading systems developers to plug-in their own technical indicators / trading rules, performance engines and adaptation routines.

One of the strongest features of the Dakota application is that it functions on a 100% walk-forward basis. When price data for given trading period (day) is processed, trading system parameters are adapted using data up to the most recent period. Whether you are building new systems or updating existing systems Dakota runs in 100% walk-forward mode. It is my belief that eventually all trading systems software will operate on this basis.

Some trading systems software applications run walk-forward simulations by organizing in-sample and out-of-sample periods so that a new out-of-sample period starts when the prior out-of-sample period finishes. This is the approach that we used over 10 years ago. The 'block by block' approach is better than optimizing over the entire simulation period and thereby producing no out-of-sample results. However, the fact that historical optimizations occur over distinct and potentially very different periods introduces numerous issues that are difficult to work around. Dakota is superior because historical optimizations are occurring bar by bar. i.e. There is not a huge 'jump' from one optimization period to the next. This enables smooth adjustments in trading system parameters to take place walking-forward.

A Dakota trading system is made up of a set of trading robots. The set of bots is referred to as a swarm. Typically, each bot in a swarm is based on the same technical indicators and trading rules. Parameter values for technical indicators and trading rules vary from bot to bot within user defined ranges and are adjusted as new data is processed. Adaptation occurs via the bar by bar adjustments in bot parameter values. The trading system signal is generated by combining the trading signals generated by the bots in the swarm.

The swarm adaptation engine is responsible for the bar by bar adaptation of bot parameter values. A trading systems developer can use any technology or set of rules to determine how the adjustments are calculated. For example, a given trade bot's parameter values can be moved closer to those of the best performing bot in the swarm as well as the trade bot's best position over the performance lookback period. Any suitable algorithm can be used by the trading systems developer to enable the adaptation process.

Trading system parameters are generally adjusted gradually period by period by the swarm adaptation engine. Although, relatively rapid adjustments in parameter values occur when significant changes in market behaviour takes place. For example, a simple moving average is used along with some trading rules by each of the 60 bots in the swarm. The length of the SMA is set to vary anywhere from 2 to 30 trading days. Bot number 1 is currently using a 9 period SMA and bot number 2 is currently using a 12 period SMA and so on. The minimum period used by the bots in the swarm is 8 trading days and the maximum is 15 trading days. In 20 trading days time the period of the SMA used by each bot will have potentially changed due to the bar by bar adaptation that is taking place. The minimum period is now 11 trading days and the maximum is 18 trading days.

The spread in parameter values provided by the swarm of bots is just as valuable as the adaptation process itself. The bots in a swarm will tend to be in the vicinity of the historically best performing bot. The best set of parameter values in the future is often not equal to the parameter values of the currently best performing bot, but will often be in the vicinity of those parameter values. Trading simulations that exclusively use the signal generated by the best performing bot in the swarm always produce inferior results compared to using the average of all bot signals in the swarm.

If any points are not clear then let me know and I'll do my best to provide a better explanation.

 

[Pat494]

Thanks James

unfortunately my disc drive collapsed recently and had to be replaced so I no longer have any Biocomp programmes