3 ms·
This is mostly a list counting how popular certain topics are, with lots of references. I am missing any actual 'review' about which methods had more success,
by rocqua 7y ago
This is mostly a list counting how popular certain topics are, with lots of references.
I am missing any actual 'review' about which methods had more success, which methods show promise, and which 'subjects' seem more amenable to Deep Learning that others.
- scottlocklin 7y agoSeveral of the papers listed there are known (by me anyway) to be complete bullshit. This seems to be "here are search results on google scholar." I actually don't know anyone using deep learning the hedge fund business, other than for screwing around. It's a terrible tool for that sort of thing. And as someone pointed out below; predicting the future is only a small part of what a trading strategy is (for some trading strategies, forecasting is actually the null set).
- chasedehan 7y agoThis is an academic literature review article. The point of these is to basically list all of the relevant papers, not discuss the relative merits of each. They are designed as a starting point for academic research.