Is there any lab work on topic such as deep learning + financial trading?

[doridori]

Hi I'm CS undergrad with 1 year industry experience, 1+ years of lab intern experience (machine learning, computational biology). I'm trying to apply at phD programs. 

I love machine learning but during my internship, I found I'm not that really interested in computational biology as I first thought and I want to search labs where I can research deep learning and trading system such as stock recommendation system, trading agent using RL, stock forecasting with various deep models, interpreting chart etc...

I've been told at ML courses  RL and machine learning system has been successful at these fields but I couldn't find many papers or labs doing deep learning and financial trading. 

Is there any university or labs with these research topic? I want to study with researchers with same interest and look for labs, but don't know where to start

[Arkhy]

Unfortunately, trading companies are not eager to share their work (for obvious reasons). The ones that do get published are usually laughable.

TBH, judging from my short experience in a hedge fund that tried to do that -- DL just might not be a good fit.  I suspect that companies that do claim success in do it because of the deep learning hype.

But don't take my word on it -- I'm not a specialist in trading and my experience took time a few years back. 

A simple search now gave me much more results  -- like https://arxiv.org/pdf/1602.06561v2.pdf , you can look up the authors. But I'd still expect the research to happen in industry labs, not in academia.

[Elephas]

You can look up the Deep Learning in Finance Summit and look at who has spoken to find professors and universities for example I saw listed:  University of Oxford, Hanken School of Economics, and they also listed some labs and companies interested in these topics.

This is a youtube video from the summit that seems aligned with your research interests: 

 

[Luptior]

http://csrankings.org/#/index?ecom

Guess in those professors who publish in comp econ should be some we do ML+FE