Help us understand the problem. What is going on with this article?

FX system trading code with Deep Learning on Python

wrote an FX system trading program with deep learning method.
I used Keras library for this deep learning implementation.

trends of portfolio

USDJPY, on Year 2000-2008, spread=0.03Yen(3pips), interval of exchange data=5min
dl.png

source code

https://github.com/ryogrid/fx_systrade/blob/3cc5975f77893c8fd50d8c00bc618916fe0b439c/keras_trade.py

strategy

  1. input some technical index, past several prices, chart type (I defined) as features.
  2. predict UP or Down with input features through deep learning.
  3. with predicted result, get position. then call off position on my original rule (please read code :-)).

if you have any question, please mail me.
ryo.contact [at] gmail.com

in Japanese:
http://qiita.com/ryo_grid/items/7746528f8cae8026b936
http://qiita.com/ryo_grid/items/56a086fee1b780157df7

just for reference

With Xgboost library (boosted decision tree), I got result below.
xgboost.png

https://github.com/ryogrid/fx_systrade/blob/a2a905fcfeb6d1e1ee91e71e0005ca17cc94757f/xgboost_trade.py

below code can real and demo trade on OANDA through REST trade API.
https://github.com/ryogrid/fx_systrade/blob/master/xgboost_trade_oanda.py

ryo_grid
svengali,kikker,はてブおせっかい,TOFU-G,ひらめいったー,キャチボ,顔面偏差値スカウター, UZOMUZO等の作者.機械学習でシステムトレードしたりも.元の専門はHPC (修士).機械学習,分散システム, physical computing,仮想通貨, 投資とか好き.Scrum-Upプロジェクト参画中. ryo.contact[at]gmail.com
https://ryogrid.net
orylab
オリィ研究所は距離や身体的問題を克服し、あいたい人に会え、行きたいところに行ける未来を実現します。
http://orylab.com
Why not register and get more from Qiita?
  1. We will deliver articles that match you
    By following users and tags, you can catch up information on technical fields that you are interested in as a whole
  2. you can read useful information later efficiently
    By "stocking" the articles you like, you can search right away