Informed Machine Learning - A Taxonomy and Survey of Integrating Prior Knowledge into Learning Systems

Rueden, Laura von; Mayer, Sebastian; Beckh, Katharina; Georgiev, Bogdan; Giesselbach, Sven; Heese, Raoul; Kirsch, Birgit; Pfrommer, Julius; Pick, Annika; Ramamurthy, Rajkumar; Walczak, Michał; Garcke, Jochen; Bauckhage, Christian; Schuecker, Jannis

doi:10.1109/tkde.2021.3079836

Cited by 346 publications

(264 citation statements)

References 133 publications

(296 reference statements)

Supporting

Mentioning

210

Contrasting

Unclassified

Order By: Relevance

“…Recent developments of machine learning algorithms applied in fluid mechanics point to a very important line of research where we ought not to entirely rely on a brute-force strategy when designing and applying a machine learning algorithm in a flow problem, but should rather consider embedding some of the most fundamental physical or mathematical constraints or using some prior knowledge in the construction of the algorithm. This idea is drawing increasing attention in the broad fields of physics and engineering (von Rueden et al 2021) as it will significantly reduce the searching space or guide the algorithm to advance in a more physically relevant direction, which will help to converge to the sought solutions more rapidly. For example, in the work of Ling, Kurzawski & Templeton (2016), a deep learning approach to RANS turbulence modelling that embedded Galilean invariance into the network using a higher-order multiplicative layer was presented.…”

Section: Reinforcement Learning As a Flow Control Strategymentioning

confidence: 99%

Reinforcement-learning-based control of confined cylinder wakes with stability analyses

Zhang

2021

J. Fluid Mech.

View full text Add to dashboard Cite

This work studies the application of a reinforcement learning (RL)-based flow control strategy to the flow past a cylinder confined between two walls to suppress vortex shedding. The control action is blowing and suction of two synthetic jets on the cylinder. The theme of this study is to investigate how to use and embed physical information of the flow in the RL-based control. First, global linear stability and sensitivity analyses based on the time-mean flow and the steady flow (which is a solution to the Navier–Stokes equations) are conducted in a range of blockage ratios and Reynolds numbers. It is found that the most sensitive region in the wake extends itself when either parameter increases in the parameter range we investigated here. Then, we use these physical results to help design RL-based control policies. We find that the controlled wake converges to the unstable steady base flow, where the vortex shedding can be successfully suppressed. A persistent oscillating control seems necessary to maintain this unstable state. The RL algorithm is able to outperform a gradient-based optimisation method (optimised in a certain period of time) in the long run. Furthermore, when the flow stability information is embedded in the reward function to penalise the instability, the controlled flow may become more stable. Finally, according to the sensitivity analyses, the control is most efficient when the probes are placed in the most sensitive region. The control can be successful even when few probes are properly placed in this manner.

show abstract

Section: Reinforcement Learning As a Flow Control Strategymentioning

confidence: 99%

Reinforcement-learning-based control of confined cylinder wakes with stability analyses

Zhang

2021

J. Fluid Mech.

View full text Add to dashboard Cite

show abstract

“…While traditional ML develops methods to derive models purely from data, more and more researchers call for combining data-and knowledge-based approaches, which can reduce the required amount of training data and, at the same time, lead to better model quality, explainability, and trustworthiness. A research field called informed ML [16] works on integrating machine learning techniques with processing of conceptual and contextual knowledge. Researchers mostly focus on utilizing knowledge that has been previously prepared and represented in a machine-readable form, such as logic rules, algebraic equations, or concept graphs.…”

Section: Informed MLmentioning

confidence: 99%

Visual Analytics for Human-Centered Machine Learning

Andrienko

Adilova

Wrobel

et al. 2022

IEEE Comput. Grap. Appl.

View full text Add to dashboard Cite

We introduce a new research area in Visual Analytics (VA) aiming to bridge existing gaps between methods of interactive Machine Learning (ML) and eXplainable Artificial Intelligence (XAI), on one side, and human minds, on the other side. The gaps are, first, a conceptual mismatch between ML/XAI outputs and human mental models and ways of reasoning, second, a mismatch between the information quantity and level of detail and human capabilities to perceive and understand. A grand challenge is to adapt ML and XAI to human goals, concepts, values, and ways of thinking. Complementing the current efforts in XAI towards solving this challenge, VA can contribute by exploiting the potential of visualization as an effective way of communicating information to humans and a strong trigger of human abstractive perception and thinking. We propose a cross-disciplinary research framework and formulate research directions for VA. THE IMPORTANCE of involving humans in the process of creating and training Machine Learning (ML) models is currently widely recognized in the ML community [1]. It is argued that humans involved in this process need to understand what the machine is doing and how it uses human inputs; hence, the machine must be able to explain its behavior to the users. Understanding of ML models has also critical importance for deciding whether they can be adopted for practical use. Explainability of models may even be more important than their performance, especially in high-stake domains. In response to the need to explain untransparent ML models ("black boxes") to users, the research field of eXplainable Artificial Intelligence (XAI) has emerged recently [8]. The work in this field was boosted by the European Parliament's adoption of the General Data Protection Regulation (GDPR), which introduces the right of

show abstract

“…Thus, their attention is close to 0, which leads to almost no information propagation between nodes of different types. This extension of GMN's original cross-graph propagation component can be seen as a form of informed machine learning [51], where prior knowledge is used within a machine learning setup such as our sGMN. See our previous work [22] for a discussion and an application scenario of informed machine learning in POCBR.…”

Section: Constrained Propagation In the Semantic Graph Matching Networkmentioning

confidence: 99%

Using Graph Embedding Techniques in Process-Oriented Case-Based Reasoning

Hoffmann

Bergmann

2022

Algorithms

View full text Add to dashboard Cite

Similarity-based retrieval of semantic graphs is a core task of Process-Oriented Case-Based Reasoning (POCBR) with applications in real-world scenarios, e.g., in smart manufacturing. The involved similarity computation is usually complex and time-consuming, as it requires some kind of inexact graph matching. To tackle these problems, we present an approach to modeling similarity measures based on embedding semantic graphs via Graph Neural Network (GNNs). Therefore, we first examine how arbitrary semantic graphs, including node and edge types and their knowledge-rich semantic annotations, can be encoded in a numeric format that is usable by GNNs. Given this, the architecture of two generic graph embedding models from the literature is adapted to enable their usage as a similarity measure for similarity-based retrieval. Thereby, one of the two models is more optimized towards fast similarity prediction, while the other model is optimized towards knowledge-intensive, more expressive predictions. The evaluation examines the quality and performance of these models in preselecting retrieval candidates and in approximating the ground-truth similarities of a graph-matching-based similarity measure for two semantic graph domains. The results show the great potential of the approach for use in a retrieval scenario, either as a preselection model or as an approximation of a graph similarity measure.

show abstract

Informed Machine Learning - A Taxonomy and Survey of Integrating Prior Knowledge into Learning Systems

Cited by 346 publications

References 133 publications

Reinforcement-learning-based control of confined cylinder wakes with stability analyses

Reinforcement-learning-based control of confined cylinder wakes with stability analyses

Visual Analytics for Human-Centered Machine Learning

Using Graph Embedding Techniques in Process-Oriented Case-Based Reasoning

Contact Info

Product

Resources

About