Hubert Baniecki

find me @hbaniecki

I’m a MSc student in Data Science at Warsaw University of Technology; working in a machine learning research lab lead by Przemysław Biecek. My research interests include explainable machine learning, responsibility, robustness, and adversarial attacks. I’m also developing open-source Python & R packages for explainability of predictive models.

news

May 2021	Contributing to the useR! 2021 conference with a talk Open the Machine Learning Black-Box with modelStudio & Arena, and through a workshop Introduction to Responsible Machine Learning.
Feb 2021	Started my MSc in Data Science at Warsaw University of Technology.
Feb 2021	Bachelor of Engineering in Data Science Obtained a Bachelor’s degree in Data Science at Faculty of Mathematics and Information Science, Warsaw University of Technology with thesis entitled Adversarial attacks on Explainable AI methods.
Nov 2020 new paper	Responsible Prediction Making of COVID-19 Mortality (Student Abstract) To be published in the proceedings of the AAAI Conference on Artificial Intelligence (AAAI).

recent papers

preprint

Fooling Partial Dependence via Data Poisoning

Baniecki, H., Kretowicz, W., and Biecek, P.

arXiv:2105.12837 2021

Abs arXiv PDF Code

Many methods have been developed to understand complex predictive models and high expectations are placed on post-hoc model explainability. It turns out that such explanations are not robust nor trustworthy, and they can be fooled. This paper presents techniques for attacking Partial Dependence (plots, profiles, PDP), which are among the most popular methods of explaining any predictive model trained on tabular data. We showcase that PD can be manipulated in an adversarial manner, which is alarming, especially in financial or medical applications where auditability became a must-have trait supporting black-box models. The fooling is performed via poisoning the data to bend and shift explanations in the desired direction using genetic and gradient algorithms. To the best of our knowledge, this is the first work performing attacks on variable dependence explanations. The novel approach of using a genetic algorithm for doing so is highly transferable as it generalizes both ways: in a model-agnostic and an explanation-agnostic manner.
preprint

The Grammar of Interactive Explanatory Model Analysis

Baniecki, H., and Biecek, P.

arXiv:2005.00497 2021

Abs arXiv PDF Website

The growing need for in-depth analysis of predictive models leads to a series of new methods for explaining their local and global properties. Which of these methods is the best? It turns out that this is an ill-posed question. One cannot sufficiently explain a black-box machine learning model using a single method that gives only one perspective. Isolated explanations are prone to misunderstanding, which inevitably leads to wrong or simplistic reasoning. This problem is known as the Rashomon effect and refers to diverse, even contradictory interpretations of the same phenomenon. Surprisingly, the majority of methods developed for explainable machine learning focus on a single aspect of the model behavior. In contrast, we showcase the problem of explainability as an interactive and sequential analysis of a model. This paper presents how different Explanatory Model Analysis (EMA) methods complement each other and why it is essential to juxtapose them together. The introduced process of Interactive EMA (IEMA) derives from the algorithmic side of explainable machine learning and aims to embrace ideas developed in cognitive sciences. We formalize the grammar of IEMA to describe potential human-model dialogues. IEMA is implemented in the human-centered framework that adopts interactivity, customizability and automation as its main traits. Combined, these methods enhance the responsible approach to predictive modeling.
AAAI

Responsible Prediction Making of COVID-19 Mortality (Student Abstract)

Baniecki, H., and Biecek, P.

AAAI Conference on Artificial Intelligence (AAAI) 2021

Abs PDF Poster Website

For high-stakes prediction making, the Responsible Artificial Intelligence (RAI) is more important than ever. It builds upon Explainable Artificial Intelligence (XAI) to advance the efforts in providing fairness, model explainability, and accountability to the AI systems. During the literature review of COVID-19 related prognosis and diagnosis, we found out that most of the predictive models are not faithful to the RAI principles, which can lead to biassed results and wrong reasoning. To solve this problem, we show how novel XAI techniques boost transparency, reproducibility and quality of models.