Human computation applied to automated trading; research project now live
Hi all, Just wanted to share some informations of a trading research project named Krabott that began in 2009 in a french university lab.
... Would be happy to share ideas and knowledge with anyone interested in this field !
Several years ago I had an idea where I imagine a boy playing a race car video game that was connected in some way to the price movements of the market, i.e., to win the race he had to beat the market in somehow. My vision was that this would be someday a reality to extract human intelligence in a different way to solve trading complex work.
Reading your ideas and PhD line of study I see a lot of synergy about this old vision and what we can expect in reality.
So, in this sense, can you share some experiences and expectations, since we will see this in action just next month? Thanks.
arnovinc, good idea, to be honest I like too much of such quantitative approaches to create/discover automated trading solutions, so count on me to share ideas.
Several years ago I had an idea where I imagine a boy playing a race car video game that was connected in some way to the price movements of the market, i.e., to win the race he had to beat the market in somehow. My vision was that this would be someday a reality to extract human intelligence in a different way to solve trading complex work.
Reading your ideas and PhD line of study I see a lot of synergy about this old vision and what we can expect in reality.
So, in this sense, can you share some experiences and expectations, since we will see this in action just next month? Thanks.
Thanks for your comment, race car trading is a great idea !!! I had the same some years ago to turn high frequency trading into a kind of bobsleigh 3D game (I think I still have a small demo somewhere...) Idea was to travel between 2 mountains with a design that follows in real time market books : buyers on the left, sellers on the right and you have to trade with the Joystick (I think you become mad in one hour). Could be based on the game Tux Racer for example (as it is open source)
I just thought again about this idea by testing those 3D glasses http://www.oculusvr.com/ few weeks ago, imagine you can play inside the market itself for high frequency scalping.... (I think you become mad in less than half an hour then).
Krabott trading signal will be provided on MQL5.com signal market place in some weeks (we are currently testing it to check all interfaces). So we will communicate in less than one month on this system.
Thanks for your work as moderator.
Thanks for your comment, race car trading is a great idea !!! I had the same some years ago to turn high frequency trading into a kind of bobsleigh 3D game (I think I still have a small demo somewhere...) Idea was to travel between 2 mountains with a design that follows in real time market books : buyers on the left, sellers on the right and you have to trade with the Joystick (I think you become mad in one hour). Could be based on the game Tux Racer for example (as it is open source)
I just thought again about this idea by testing those 3D glasses http://www.oculusvr.com/ few weeks ago, imagine you can play inside the market itself for high frequency scalping.... (I think you become mad in less than half an hour then).
Krabott trading signal will be provided on MQL5.com signal market place in some weeks (we are currently testing it to check all interfaces). So we will communicate in less than one month on this system.
Thanks for your work as moderator.
Great, so we are on the same page, it would be a nice way to win money playing such game ;-)
You are welcome.
I always appreciate every person who has a "beyond" the vision, thoughts and ideas. Although it does not always deliver the results we expect, to conquer the market. Although I am not a fan of the game, but I think this is one good approach as stated by figurelli.
Good job arnovinc and good luck..
I always appreciate every person who has a "beyond" the vision, thoughts and ideas. Although it does not always deliver the results we expect, to conquer the market. Although I am not a fan of the game, but I think this is one good approach as stated by figurelli.
Good job arnovinc and good luck..
Thanks ! My thesis will be published in some weeks on the net, I'll put a link in this forum. But unfortunately in French only
French is just good
Indeed, by the way it must be in the plans, if MQ use this kind of statistics: List of languages by number of native speakers
Anyway, a crazy idea would be a free idioms area in the Forum, can you imagine moderate such area? ;-)
Indeed, by the way it must be in the plans, if MQ use this kind of statistics: List of languages by number of native speakers
Anyway, a crazy idea would be a free idioms area in the Forum, can you imagine moderate such area? ;-)
Thank you for the link, I was just thinking about these stats.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hi all, Just wanted to share some informations of a trading research project named Krabott that began in 2009 in a french university lab.
Maybe you heard about Fold-it experience in biochemistery . The Fold it project predicts the 3D shape of proteins given their amino acid sequence, using a game. Teams of on-line volunteers with no special knowledge who were invited to play a collective game came up with better solutions to the puzzles than state of the art computer programs. Some people call this a "human computation" process. (see http://fold.it)
Since 2009, I planed to apply the same recipes to automated trading. The aim is to ask players without any financial or technical knowledge, to design new parameters of existing trading strategies. Name of the project is "Krabott". Calculation engine is based on genetic algorithmes HBGA (described by a researcher A.Kosorukoff in 2000). (see http://www.krabott.com)
After 4 years and a PhD on this subject we now are ready to publish as a signal provider a "crowdcomputed" trading signal, only based on work of severals players, thanks to MQL5.com community this may be ready before the end of february. The signal is not an average opinion of several players, but the result of a trading strategy where all parameters are computed by the crowd.
Would be happy to share ideas and knowledge with anyone interested in this field !