Additive Tree Ensembles: Reasoning About Potential Instances

by   Laurens Devos, et al.

Imagine being able to ask questions to a black box model such as "Which adversarial examples exist?", "Does a specific attribute have a disproportionate effect on the model's prediction?" or "What kind of predictions are possible for a partially described example?" This last question is particularly important if your partial description does not correspond to any observed example in your data, as it provides insight into how the model will extrapolate to unseen data. These capabilities would be extremely helpful as it would allow a user to better understand the model's behavior, particularly as it relates to issues such as robustness, fairness, and bias. In this paper, we propose such an approach for an ensemble of trees. Since, in general, this task is intractable we present a strategy that (1) can prune part of the input space given the question asked to simplify the problem; and (2) follows a divide and conquer approach that is incremental and can always return some answers and indicates which parts of the input domains are still uncertain. The usefulness of our approach is shown on a diverse set of use cases.


page 1

page 2

page 3

page 4


Adversarial Example Detection in Deployed Tree Ensembles

Tree ensembles are powerful models that are widely used. However, they a...

An Efficient Adversarial Attack for Tree Ensembles

We study the problem of efficient adversarial attacks on tree based ense...

On the Perils of Cascading Robust Classifiers

Ensembling certifiably robust neural networks has been shown to be a pro...

Learning User-Interpretable Descriptions of Black-Box AI System Capabilities

Several approaches have been developed to answer specific questions that...

Robustness to Adversarial Examples through an Ensemble of Specialists

We are proposing to use an ensemble of diverse specialists, where specia...

Imagining the Unseen: Learning a Distribution over Incomplete Images with Dense Latent Trees

Images are composed as a hierarchy of object parts. We use this insight ...