Using Recurrent Neural Networks To Forecasting of Forex — mentioned by dsacco

Thu, 08 Sep 2016 21:47:29 GMT

To your first question: yes, for some definitions of "successfully"[1][2][3][4]. It's difficult, and it isn't as though you train your algorithm and watch it make money. It's more that some functions of data analysis become incrementally easier or superior. It's fundamentally similar to applying RNNs to any other signal processing problem. To your (implied) second question: tech companies like Google, Facebook and Microsoft have a strong incentive to publish their researchers' work. It's good for recruiting and pushing progress forward overall. Financial companies have a strong incentive not…

Paper and discussion history. A mention is not an endorsement.

Using Recurrent Neural Networks To Forecasting of Forex · HackerPapers discussions

Using Recurrent Neural Networks To Forecasting of Forex — mentioned by dsacco