A5015971884- Fluid Dynamics and Turbulent Flows
- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- Fluid Dynamics and Vibration Analysis
- Bayesian Modeling and Causal Inference
- Plant Water Relations and Carbon Dynamics
- Anomaly Detection Techniques and Applications
- Machine Learning and Data Classification
- Computational Physics and Python Applications
- Gaussian Processes and Bayesian Inference
- Time Series Analysis and Forecasting
- Innovative Educational Technologies
- Computational Drug Discovery Methods
- Machine Learning in Materials Science
- Model Reduction and Neural Networks
- Gene expression and cancer classification
German Center for Neurodegenerative Diseases
2021-2023
Abstract Causal learning is a key challenge in scientific artificial intelligence as it allows researchers to go beyond purely correlative or predictive analyses towards underlying cause-and-effect relationships, which are important for understanding well wide range of downstream tasks. Here, motivated by emerging biomedical questions, we propose deep neural architecture causal relationships between variables from combination high-dimensional data and prior knowledge. We combine...
Abstract Particle-image velocimetry (PIV) is one of the key techniques in modern experimental fluid mechanics to determine velocity components flow fields a wide range complex engineering problems. Current PIV processing tools are mainly handcrafted models based on cross-correlations computed across interrogation windows. Although widely used, these existing have number well-known shortcomings, including limited spatial output resolution and peak-locking biases. Recently, new approaches for...
We propose a method for learning dynamical systems from high-dimensional empirical data that combines variational autoencoders and (spatio-)temporal attention within framework designed to enforce certain scientifically-motivated invariances. focus on the setting in which are available multiple different instances of system whose underlying model is entirely unknown at outset. The approach rests separation into an instance-specific encoding (capturing initial conditions, constants etc.)...
Recent years have seen rapid progress at the intersection between causality and machine learning. Motivated by scientific applications involving high-dimensional data, in particular biomedicine, we propose a deep neural architecture for learning causal relationships variables from combination of empirical data prior knowledge. We combine convolutional graph networks within risk framework to provide flexible scalable approach. Empirical results include linear nonlinear simulations (where...