Lenon Minorics — accepted AI-conference papers

2024

Quantifying intrinsic causal contributions via structure preserving interventions

Dominik Janzing, Patrick Blöbaum, Atalanti A Mastakouri, Philipp M Faller, Lenon Minorics, Kailash Budhathoki

AISTATS 2024poster

We propose a notion of causal influence that describes the ‘intrinsic’ part of the contribution of a node on a target node in a DAG. By recursively writing each node as a function of the upstream noise terms, we separate the intrinsic information added by each node from the one obtained from its anc…

Cited by 11 Source ↗PDF ↗Scholar ↗

2023

Manifold Restricted Interventional Shapley Values

Muhammad Faaiz Taufiq, Patrick Blöbaum, Lenon Minorics

AISTATS 2023poster

Shapley values are model-agnostic methods for explaining model predictions. Many commonly used methods of computing Shapley values, known as off-manifold methods, rely on model evaluations on out-of-distribution input samples. Consequently, explanations obtained are sensitive to model behaviour outs…

Cited by 11 Source ↗PDF ↗code ↗Scholar ↗

2023

Unsupervised Model Selection for Time Series Anomaly Detection

Mononito Goswami, Cristian Ignacio Challu, Laurent Callot, Lenon Minorics, Andrey Kan

ICLR 2023top-25%

Anomaly detection in time-series has a wide range of practical applications. While numerous anomaly detection methods have been proposed in the literature, a recent survey concluded that no single method is the most accurate across various datasets. To make matters worse, anomaly labels are scarce a…

Cited by 52 Source ↗PDF ↗code ↗Scholar ↗

2022

Causal forecasting: generalization bounds for autoregressive models

Leena Chennuru Vankadara, Philipp Michael Faller, Michaela Hardt, Lenon Minorics, Debarghya Ghoshdastidar, Dominik Janzing

UAI 2022poster

Despite the increasing relevance of forecasting methods, causal implications of these algorithms remain largely unexplored. This is concerning considering that, even under simplifying assumptions such as causal sufficiency, the statistical risk of a model can differ significantly from its causal ris…

Cited by 9 Source ↗PDF ↗code ↗Scholar ↗

2022

Causal structure-based root cause analysis of outliers

Kailash Budhathoki, Lenon Minorics, Patrick Bloebaum, Dominik Janzing

ICML 2022spotlight

Current techniques for explaining outliers cannot tell what caused the outliers. We present a formal method to identify "root causes" of outliers, amongst variables. The method requires a causal graph of the variables along with the functional causal model. It quantifies the contribution of each var…

Cited by 71 Source ↗PDF ↗code ↗Scholar ↗

2022

Testing Granger Non-Causality in Panels with Cross-Sectional Dependencies

Lenon Minorics, Caner Turkmen, David Kernert, Patrick Bloebaum, Laurent Callot, Dominik Janzing

AISTATS 2022poster

This paper proposes a new approach for testing Granger non-causality on panel data. Instead of aggregating panel member statistics, we aggregate their corresponding p-values and show that the resulting p-value approximately bounds the type I error by the chosen significance level even if the panel m…

Cited by 2 Source ↗PDF ↗code ↗Scholar ↗

2020

Feature relevance quantification in explainable AI: A causal problem

Dominik Janzing, Lenon Minorics, Patrick Bloebaum

AISTATS 2020poster

We discuss promising recent contributions on quantifying feature relevance using Shapley values, where we observed some confusion on which probability distribution is the right one for dropped features. We argue that the confusion is based on not carefully distinguishing between observational and in…

Cited by 429 Source ↗PDF ↗Scholar ↗