4 ms·
Interactive Brokers has paper trading and good docs http://www.interactivebrokers.com/en/p.php?f=tws&p=p http://www.interactivebrokers.com/en/p.php?f=tws&p=p ht
by arthurdent 16y ago
Interactive Brokers has paper trading and good docs
http://www.interactivebrokers.com/en/p.php?f=tws&p=p http://www.interactivebrokers.com/en/p.php?f=tws&p=p
http://www.interactivebrokers.com/en/p.php?f=programInterface http://www.interactivebrokers.com/en/p.php?f=programInterfac...
http://www.maxdama.com/ http://www.maxdama.com/ talks about the IB api a bit.
I skimmed your comments and kept seeing the word python. There's a 3rd party python-IB package here (can't vouch for it):
http://code.google.com/p/ibpy/ http://code.google.com/p/ibpy/
http://epchan.blogspot.com/ http://epchan.blogspot.com/ talks a little bit about basic pairs trading strategies. My experience has been that adding complexity off the bat is usually not the way to go. I don't personally know anyone who is using AI/machine learning as their primary strategy generation mechanism, but maybe I just run with the wrong circles.
I'm interested in collaborating with smart people re trading ideas, so if you're interested as well, fire me an email.
- jchonphoenix 16y agoI've heard of IB before so its a name I'll probably look into. I'm mostly interested in machine learning because I feel like a lot of the algorithms I'm applying to recommendation systems could be applied to stocks. If you think of the stock market as a giant system where someone wants to recommend stocks that will go up, it appears these ML techniques may be applicable. At this point however I don't have a way to model the stock market as an acceptable input to the most promising algo I have, which is one that I developed myself and has and extremely low RMSE for certain types of data.