Moritz Grosse-Wentrup’s research while affiliated with University of Vienna and other places

Objective. To date, a comprehensive comparison of Riemannian decoding methods with deep convolutional neural networks for EEG-based Brain-Computer Interfaces remains absent from published work. We address this research gap by using MOABB, The Mother Of All BCI Benchmarks, to compare novel convolutional neural networks to state-of-the-art Riemannian approaches across a broad range of EEG datasets, including motor imagery, P300, and steady-state visual evoked potentials paradigms. Approach. We systematically evaluated the performance of convolutional neural networks, specifically EEGNet, shallow ConvNet, and deep ConvNet, against well-established Riemannian decoding methods using MOABB processing pipelines. This evaluation included within- session, cross-session, and cross-subject methods, to provide a practical analysis of model effectiveness and to find an overall solution that performs well across different experimental settings. Main results. We find no significant differences in decoding performance between convolutional neural networks and Riemannian methods for within-session, cross-session, and cross-subject analyses. Significance. The results show that, when using traditional Brain-Computer Interface paradigms, the choice between CNNs and Riemannian methods may not heavily impact decoding performances in many experimental settings. These findings provide researchers with flexibility in choosing decoding approaches based on factors such as ease of implementation, computational efficiency or individual preferences.

Explaining how neuronal activity gives rise to cognition arguably remains the most significant challenge in cognitive neuroscience. We introduce neuro-cognitive multilevel causal modeling (NC-MCM), a framework that bridges the explanatory gap between neuronal activity and cognition by construing cognitive states as (behaviorally and dynamically) causally consistent abstractions of neuronal states. Multilevel causal modeling allows us to interchangeably reason about the neuronal- and cognitive causes of behavior while maintaining a physicalist (in contrast to a strong dualist) position. We introduce an algorithm for learning cognitive-level causal models from neuronal activation patterns and demonstrate its ability to learn cognitive states of the nematode C. elegans from calcium imaging data. We show that the cognitive-level model of the NC-MCM framework provides a concise representation of the neuronal manifold of C. elegans and its relation to behavior as a graph, which, in contrast to other neuronal manifold learning algorithms, supports causal reasoning. We conclude the article by arguing that the ability of the NC-MCM framework to learn causally interpretable abstractions of neuronal dynamics and their relation to behavior in a purely data-driven fashion is essential for understanding more biological systems whose complexity prohibits the development of hand-crafted computational models.

Neuronal manifold learning techniques represent high-dimensional neuronal dynamics in low-dimensional embeddings to reveal the intrinsic structure of neuronal manifolds. Common to these techniques is their goal to learn low-dimensional embeddings that preserve all dynamic information in the high-dimensional neuronal data, i.e., embeddings that allow for reconstructing the original data. We introduce a novel neuronal manifold learning technique, BunDLe-Net, that learns a low-dimensional Markovian embedding of the neuronal dynamics which preserves only those aspects of the neuronal dynamics that are relevant for a given behavioural context. In this way, BunDLe-Net eliminates neuronal dynamics that are irrelevant to decoding behaviour, effectively de-noising the data to reveal better the intricate relationships between neuronal dynamics and behaviour. We demonstrate the quantitative superiority of BunDLe-Net over commonly used and state-of-the-art neuronal manifold learning techniques in terms of dynamic and behavioural information in the learned manifold on calcium imaging data recorded in the nematode C. elegans. Qualitatively, we show that BunDLe-Net learns highly consistent manifolds across multiple worms that reveal the neuronal and behavioural motifs that form the building blocks of the neuronal manifold.

Algorithmic recourse recommendations inform stakeholders of how to act to revert unfavorable decisions. However, existing methods may recommend actions that lead to acceptance (i.e., revert the model's decision) but do not lead to improvement (i.e., may not revert the underlying real-world state). To recommend such actions is to recommend fooling the predictor. We introduce a novel method, Improvement-Focused Causal Recourse (ICR), which involves a conceptual shift: Firstly, we require ICR recommendations to guide toward improvement. Secondly, we do not tailor the recommendations to be accepted by a specific predictor. Instead, we leverage causal knowledge to design decision systems that predict accurately pre- and post-recourse, such that improvement guarantees translate into acceptance guarantees. Curiously, optimal pre-recourse classifiers are robust to ICR actions and thus suitable post-recourse. In semi-synthetic experiments, we demonstrate that given correct causal knowledge ICR, in contrast to existing approaches, guides toward both acceptance and improvement.

The Shapley Additive Global Importance (SAGE) value is a theoretically appealing interpretability method that fairly attributes global importance to a model's features. However, its exact calculation requires the computation of the feature's surplus performance contributions over an exponential number of feature sets. This is computationally expensive, particularly because estimating the surplus contributions requires sampling from conditional distributions. Thus, SAGE approximation algorithms only take a fraction of the feature sets into account. We propose d-SAGE, a method that accelerates SAGE approximation. d-SAGE is motivated by the observation that conditional independencies (CIs) between a feature and the model target imply zero surplus contributions, such that their computation can be skipped. To identify CIs, we leverage causal structure learning (CSL) to infer a graph that encodes (conditional) independencies in the data as d-separations. This is computationally more efficient because the expense of the one-time graph inference and the d-separation queries is negligible compared to the expense of surplus contribution evaluations. Empirically we demonstrate that d-SAGE enables the efficient and accurate estimation of SAGE values.

Algorithmic recourse recommendations, such as Karimi et al.'s (2021) causal recourse (CR), inform stakeholders of how to act to revert unfavourable decisions. However, some actions lead to acceptance (i.e., revert the model's decision) but do not lead to improvement (i.e., may not revert the underlying real-world state). To recommend such actions is to recommend fooling the predictor. We introduce a novel method, Improvement-Focused Causal Recourse (ICR), which involves a conceptual shift: Firstly, we require ICR recommendations to guide towards improvement. Secondly, we do not tailor the recommendations to be accepted by a specific predictor. Instead, we leverage causal knowledge to design decision systems that predict accurately pre- and post-recourse. As a result, improvement guarantees translate into acceptance guarantees. We demonstrate that given correct causal knowledge, ICR, in contrast to existing approaches, guides towards both acceptance and improvement.

An increasing number of model-agnostic interpretation techniques for machine learning (ML) models such as partial dependence plots (PDP), permutation feature importance (PFI) and Shapley values provide insightful model interpretations, but can lead to wrong conclusions if applied incorrectly. We highlight many general pitfalls of ML model interpretation, such as using interpretation techniques in the wrong context, interpreting models that do not generalize well, ignoring feature dependencies, interactions, uncertainty estimates and issues in high-dimensional settings, or making unjustified causal interpretations, and illustrate them with examples. We focus on pitfalls for global methods that describe the average model behavior, but many pitfalls also apply to local methods that explain individual predictions. Our paper addresses ML practitioners by raising awareness of pitfalls and identifying solutions for correct model interpretation, but also addresses ML researchers by discussing open issues for further research.