Hacker News new | past | comments | ask | show | jobs | submit login

From a personal point of view as someone also interested in learning a bit I think building something is often a good way to go. I kind of have a plan one day to hack together the neural network from

http://karpathy.github.io/2015/05/21/rnn-effectiveness/

with some personally selected stock market data to see if it can kick out profitable buy/sells.

(There is some how to stuff in the article under 'Source Code')




A "good" AI stock trader would in particular require some kind of real-world pseudo real-time data feed (pick your poison; Twitter, FB, news, even generic web crawling)

Anything based purely on the stock charts themselves becomes simply a sophisticated pattern matcher and can't possibly respond to the actual drivers of price movement.


It depends on what strategy you are trying to implement. I wasn't particularly thinking real time. I'm more of a value investor at heart and was thinking of something like it looking for stocks where the valuation is underpinned by assets and buying when they are near the bottom. So the inputs would be something like some numbers from the income and balance statements plus the price graph.


> where the valuation is underpinned by assets and buying when they are near the bottom

... is still using some data external to the stock charts, which should work great. :)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: