Hubert Baniecki

h.baniecki@uw.edu.pl

In October 2022, I will start my PhD in Computer Science at the University of Warsaw, MI² research lab led by Przemysław Biecek. I completed my MSc (2022) and BSc (2021) in Data Science at Warsaw University of Technology.

My main research interest is explainable machine learning, with particular emphasis on adversarial attacks & evaluation of explanations. I care about human-model interaction with applications in biomedicine. I support the development of several open-source Python & R packages for building predictive models responsibly.

news [archive]

2022 Sep	I got admissioned to PhD in Computer Science at University of Warsaw & with external funding from the National Science Centre (Poland) PRELUDIUM BIS research grant ARES: Attack-Resistant Explanations toward Secure and trustworthy AI. Thanks to my supervisor Przemysław Biecek.
2022 Sep	I graduated with an MSc in Data Science from Warsaw University of Technology by defending a thesis From manipulating to evaluating explanations of machine learning models.
2022 Jun	New paper Fooling Partial Dependence via Data Poisoning is accepted for publication at the ECML PKDD 2022 conference.
2022 Mar	I received the Minister’s (of Education and Science, Poland) scholarship for students for significant scientific achievements.
2022 Jan	I received the John M. Chambers Statistical Software Award for Python package dalex and got invited for a talk at the 2022 Joint Statistical Meetings, Washington DC, USA. Kudos to the dalex team!

highlights [more]

arXiv

SurvSHAP(t): Time-dependent explanations of machine learning survival models

M. Krzyziński, M. Spytek, H. Baniecki, P. Biecek

arXiv:2208.11080v2, 2022
This work introduces the first time-dependent explanation for interpreting survival models.

Abstract PDF arXiv Code

Machine and deep learning survival models demonstrate similar or even improved time-to-event prediction capabilities compared to classical statistical learning methods yet are too complex to be interpreted by humans. Several model-agnostic explanations are available to overcome this issue; however, none directly explain the survival function prediction. In this paper, we introduce SurvSHAP(t), the first time-dependent explanation that allows for interpreting survival black-box models. It is based on SHapley Additive exPlanations with solid theoretical foundations and a broad adoption among machine learning practitioners. The proposed methods aim to enhance precision diagnostics and support domain experts in making decisions. Experiments on synthetic and medical data confirm that SurvSHAP(t) can detect variables with a time-dependent effect, and its aggregation is a better determinant of the importance of variables for a prediction than SurvLIME. SurvSHAP(t) is model-agnostic and can be applied to all models with functional output. We provide an accessible implementation of time-dependent explanations in Python at https://github.com/MI2DataLab/survshap.

ECML PKDD oral

Fooling Partial Dependence via Data Poisoning

H. Baniecki, W. Kretowicz, P. Biecek

In: European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), 2022
This work is the core of our BSc thesis defended at Warsaw University of Technology in 2021.

Abstract PDF arXiv 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 using a genetic algorithm for attacking explanations, which is highly transferable as it generalizes both ways: in a model-agnostic and an explanation-agnostic manner.
JMLR award

dalex: Responsible Machine Learning with Interactive Explainability and Fairness in Python

H. Baniecki, W. Kretowicz, P. Piatyszek, J. Wisniewski, P. Biecek

Journal of Machine Learning Research, 2021
This work received the John M. Chambers Statistical Software Award by ASA in 2022.

Abstract HTML Code Website

In modern machine learning, we observe the phenomenon of opaqueness debt, which manifests itself by an increased risk of discrimination, lack of reproducibility, and deflated performance due to data drift. An increasing amount of available data and computing power results in the growing complexity of black-box predictive models. To manage these issues, good MLOps practice asks for better validation of model performance and fairness, higher explainability, and continuous monitoring. The necessity for deeper model transparency comes from both scientific and social domains and is also caused by emerging laws and regulations on artificial intelligence. To facilitate the responsible development of machine learning models, we introduce dalex, a Python package which implements a model-agnostic interface for interactive explainability and fairness. It adopts the design crafted through the development of various tools for explainable machine learning; thus, it aims at the unification of existing solutions. This library's source code and documentation are available under open license at https://python.drwhy.ai.