Danske Commodities
Trading

Set up an auto-trading team

Danske Commodities wanted to make sure it's new auto-trading machine learning team did the right things from the start. We gave the team deep insights on how to apply machine learning to trading.

Our biggest issue was that we were a fairly new team with little experience in Machine Learning. We needed some inspiration. We needed to know if we’re on the right track. If we didn’t get the input, things would not have worked out so well. A concern for us was that you get a standard walkthrough presentation that is not really related to us as a company – that they just fire of their standard slides and talks. And that was not at all the case. The feedback to the workshop was very very positive and we’re very happy that it was so concrete, that we went right to the facts – and that we had something that we could use right the following day.

Thor Kalstrup

,

Head of Automated Trading

Tools we used

Pandas
-
Great for exploration and feature engineering.
Python
-
Our main development language.
scikit-learn
-
Grab-box of algorithms and more.
XGBoost
-
Gradient boosting powerhouse.
Zipline
-
Pythonic algorithmic trading library.