Use of domain knowledge and feature engineering in helping AI to play Hearthstone

@article{Przybyszewski2017UseOD,
  title={Use of domain knowledge and feature engineering in helping AI to play Hearthstone},
  author={Przemyslaw Przybyszewski and Szymon Dziewiatkowski and Sebastian Jaszczur and Mateusz Smiech and Marcin S. Szczuka},
  journal={2017 Federated Conference on Computer Science and Information Systems (FedCSIS)},
  year={2017},
  pages={143-148}
}
This paper describes two approaches to the AAIA'17 Data Mining Challenge. Both approaches are making extensive use of domain/background knowledge about the game to build better representation of classification problem by engineering new features. With newly constructed attributes both approaches resort to Artificial Neural Networks (ANN) to construct classification model. The resulting solutions are effective and meaningful.