no code implementations • ECCV 2020 • Christian Reimers, Jakob Runge, Joachim Denzler
Deep neural networks are tremendously successful in many applications, but end-to-end trained networks often result in hard to understand black-box classifiers or predictors.
no code implementations • 7 Feb 2024 • jonas Wahl, Jakob Runge
Many state-of-the-art causal discovery methods aim to generate an output graph that encodes the graphical separation and connection statements of the causal graph that underlies the data-generating process.
no code implementations • 6 Dec 2023 • Simon Bing, jonas Wahl, Urmi Ninad, Jakob Runge
In causal models, a given mechanism is assumed to be invariant to changes of other mechanisms.
1 code implementation • 5 Nov 2023 • Simon Bing, Urmi Ninad, jonas Wahl, Jakob Runge
The task of inferring high-level causal variables from low-level observations, commonly referred to as causal representation learning, is fundamentally underconstrained.
no code implementations • 17 Oct 2023 • Oana-Iuliana Popescu, Andreas Gerhardus, Jakob Runge
One approach computes distances by a one-hot encoding of the categorical variables, essentially treating categorical variables as discrete-numerical, while the other expresses CMI by entropy terms where the categorical variables appear as conditions only.
no code implementations • 9 Oct 2023 • Andreas Gerhardus, jonas Wahl, Sofia Faltenbacher, Urmi Ninad, Jakob Runge
In this work, we develop a method for projecting infinite time series graphs with repetitive edges to marginal graphical models on a finite time window.
1 code implementation • 31 Aug 2023 • Felix Wagner, Florian Nachtigall, Lukas Franken, Nikola Milojevic-Dupont, Rafael H. M. Pereira, Nicolas Koch, Jakob Runge, Marta Gonzalez, Felix Creutzig
Here, we address all three gaps via causal graph discovery and explainable machine learning to detect urban form effects on intra-city car travel, based on mobility data of six cities across three continents.
no code implementations • 20 Jun 2023 • Wiebke Günther, Urmi Ninad, jonas Wahl, Jakob Runge
We frame heteroskedasticity in a structural causal model framework and present an adaptation of the partial correlation CI test that works well in the presence of heteroskedastic noise, given that expert knowledge about the heteroskedastic relationships is available.
1 code implementation • 15 Jun 2023 • Kevin Debeire, Jakob Runge, Andreas Gerhardus, Veronika Eyring
It can be combined with a range of time series causal discovery methods and provides a measure of confidence for the links of the time series graphs.
no code implementations • 21 May 2023 • Gustau Camps-Valls, Andreas Gerhardus, Urmi Ninad, Gherardo Varando, Georg Martius, Emili Balaguer-Ballester, Ricardo Vinuesa, Emiliano Diaz, Laure Zanna, Jakob Runge
Discovering equations, laws and principles that are invariant, robust and causal explanations of the world has been fundamental in physical sciences throughout the centuries.
1 code implementation • 11 Apr 2023 • Saranya Ganesh S., Tom Beucler, Frederick Iat-Hin Tam, Milton S. Gomez, Jakob Runge, Andreas Gerhardus
We apply our framework to the statistical intensity prediction of Western Pacific Tropical Cyclones (TC), for which it is often difficult to accurately choose drivers and their dimensionality reduction (time lags, vertical levels, and area-averaging).
2 code implementations • 16 Apr 2021 • Christian Requena-Mesa, Vitus Benson, Markus Reichstein, Jakob Runge, Joachim Denzler
We frame Earth surface forecasting as the task of predicting satellite imagery conditioned on future weather.
Ranked #5 on Earth Surface Forecasting on EarthNet2021 OOD Track
no code implementations • 10 Mar 2021 • Christian Reimers, Paul Bodesheim, Jakob Runge, Joachim Denzler
Often, the bias of a classifier is a direct consequence of a bias in the training dataset, frequently caused by the co-occurrence of relevant features and irrelevant ones.
1 code implementation • NeurIPS 2021 • Jakob Runge
In the present work optimality is characterized as maximizing a certain adjustment information which allows to derive a necessary and sufficient graphical criterion for the existence of an optimal adjustment set and a definition and algorithm to construct it.
1 code implementation • 11 Dec 2020 • Christian Requena-Mesa, Vitus Benson, Joachim Denzler, Jakob Runge, Markus Reichstein
Here, we define high-resolution Earth surface forecasting as video prediction of satellite imagery conditional on mesoscale weather forecasts.
1 code implementation • NeurIPS 2020 • Andreas Gerhardus, Jakob Runge
We show that existing causal discovery methods such as FCI and variants suffer from low recall in the autocorrelated time series case and identify low effect size of conditional independence tests as the main reason.
no code implementations • 2 Jul 2020 • Gustau Camps-Valls, Dino Sejdinovic, Jakob Runge, Markus Reichstein
Earth observation (EO) by airborne and satellite remote sensing and in-situ observations play a fundamental role in monitoring our planet.
1 code implementation • 7 Mar 2020 • Jakob Runge
The paper introduces a novel conditional independence (CI) based method for linear and nonlinear, lagged and contemporaneous causal discovery from observational time series in the causally sufficient case.
1 code implementation • 5 Sep 2017 • Jakob Runge
Combining the local permutation scheme with the kernel tests leads to better calibration, but suffers in power.
3 code implementations • 22 Feb 2017 • Jakob Runge, Dino Sejdinovic, Seth Flaxman
Detecting causal associations in time series datasets is a key challenge for novel insights into complex dynamical systems such as the Earth system or the human brain.
Methodology Atmospheric and Oceanic Physics Applications
1 code implementation • 2 Jul 2015 • Jonathan F. Donges, Jobst Heitzig, Boyan Beronov, Marc Wiedermann, Jakob Runge, Qing Yi Feng, Liubov Tupikina, Veronika Stolbova, Reik V. Donner, Norbert Marwan, Henk A. Dijkstra, Jürgen Kurths
Additionally, \texttt{pyunicorn} provides insights into the nonlinear dynamics of complex systems as recorded in uni- and multivariate time series from a non-traditional perspective by means of recurrence quantification analysis (RQA), recurrence networks, visibility graphs and construction of surrogate time series.
Data Analysis, Statistics and Probability Atmospheric and Oceanic Physics
no code implementations • 18 Jun 2015 • Jakob Runge, Reik V. Donner, Jürgen Kurths
Here a prediction scheme that overcomes this strong limitation is introduced utilizing a causal pre-selection step which drastically reduces the number of possible predictors to the most predictive set of causal drivers making a globally optimal search scheme tractable.