Rule Hiding in Decision Trees
What happens when you want to make sure that a certain node does not appear as a leaf, without having to resort to pruning?
You wouldn't want to use pruning because the original algorithm would uncover the sensitive leaf if fed with the original data set. S
o, as you don't know what algorithm your opponent will use, your best option is to "cook" the data, as little as possible.
Here are some papers:
- D. Kalles, V. Verykios and A. Papagelis. “Hiding Decision Tree Rules by Data Set Operations”, IEEE International Conference on Information, Intelligence, Systems and Applications, Corfu, Greece, July 2015.
- D. Kalles, V. Verykios, G. Feretzakis and A. Papagelis. “Data Set Operations tohide Decision Tree Rules”, International Workshop on AI for Privacy and Security, a workshop of theEuropean Conference on Artificial Intelligence, the Hague, the Netherlands, August-September, 2016.
- D. Kalles, V. Verykios, G. Feretzakis and A. Papagelis. “Data Set Operations to hide Decision Tree Rules”, European Conference on Artificial Intelligence, the Hague, the Netherlands, August-September, 2016.
- Feretzakis, G., Kalles, D., & Verykios, V. (2018). On Using Linear Diophantine Equations for Efficient Hiding of Decision Tree Rules. Proceedings Of The 10Th Hellenic Conference On Artificial Intelligence - SETN '18. doi: 10.1145/3200947.3201030
- Feretzakis, G.; Kalles, D.; Verykios, V.S. On Using Linear Diophantine Equations for in-Parallel Hiding of Decision Tree Rules. Entropy 2019, 21, 66. doi:10.3390/e21010066
Here's an online prototype (admittedly, a bit esoteric in how to use it) ...
... and a further prototype on demonstrating how the diophantine equations concept should work.