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

Abstract

Despite its great success, machine learning can have its limits when dealing with insufficient training data. A potential solution is the additional integration of prior knowledge into the training process which leads to the notion of informed machine learning. In this paper, we present a structured overview of various approaches in this field. We provide a definition and propose a concept for informed machine learning which illustrates its building blocks and distinguishes it from conventional machine learning. We introduce a taxonomy that serves as a classification framework for informed machine learning approaches. It considers the source of knowledge, its representation, and its integration into the machine learning pipeline. Based on this taxonomy, we survey related research and describe how different knowledge representations such as algebraic equations, logic rules, or simulation results can be used in learning systems. This evaluation of numerous papers on the basis of our taxonomy uncovers key methods in the field of informed machine learning.

Accepted at IEEE Transactions on Knowledge and Data Engineering: https://ieeexplore.ieee.org/document/9429985

References in corpus (41)

• Efficient Estimation of Word Representations in Vector Space
• Variational Inference: A Review for Statisticians
• Geometric deep learning: going beyond Euclidean data
• Towards A Rigorous Science of Interpretable Machine Learning
• Artificial Neural Networks for Solving Ordinary and Partial Differential Equations
• Theory-guided Data Science: A New Paradigm for Scientific Discovery from Data
• Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data
• Explainable Machine Learning for Scientific Insights and Discoveries
• Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations
• Robots that can adapt like animals
• Sim-to-Real Transfer of Robotic Control with Dynamics Randomization
• Interaction Networks for Learning about Objects, Relations and Physics
• Group Equivariant Convolutional Networks
• A Comprehensive Physics-Informed Machine Learning Framework for Predictive Turbulence Modeling
• Physics-Informed Machine Learning Approach for Augmenting Turbulence Models: A Comprehensive Framework
• Physics-guided Neural Networks (PGNN): An Application in Lake Temperature Modeling
• Deep Q-learning from Demonstrations
• Challenges of Real-World Reinforcement Learning
• Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims
• Harnessing Deep Neural Networks with Logic Rules
• Deep-learned Top Tagging with a Lorentz Layer
• Exploiting Cyclic Symmetry in Convolutional Neural Networks
• ERNIE: Enhanced Language Representation with Informative Entities
• Hinge-Loss Markov Random Fields and Probabilistic Soft Logic
• Neural-Symbolic Computing: An Effective Methodology for Principled Integration of Machine Learning and Reasoning
• Self-Supervised Intrinsic Image Decomposition
• Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning
• Simulation assisted machine learning
• Neurosymbolic AI: The 3rd Wave
• Hybrid Knowledge Routed Modules for Large-scale Object Detection
• A Semantic Loss Function for Deep Learning with Symbolic Knowledge
• Physics-informed deep generative models
• From Deep to Physics-Informed Learning of Turbulence: Diagnostics
• Beating Atari with Natural Language Guided Reinforcement Learning
• Incorporating Causal Prior Knowledge as Path-Constraints in Bayesian Networks and Maximal Ancestral Graphs
• Using Simulation to Improve Sample-Efficiency of Bayesian Optimization for Bipedal Robots
• The Good, the Bad and the Ugly: Augmenting a black-box model with expert knowledge
• Scalable Structure Learning for Probabilistic Soft Logic
• Adversarial Constraint Learning for Structured Prediction
• Compensating data shortages in manufacturing with monotonicity knowledge
• Black-Box Inference for Non-Linear Latent Force Models

Cited by in corpus (63)

• Explainable Machine Learning for Scientific Insights and Discoveries
• Machine Learning in Aerodynamic Shape Optimization
• Autonomous discovery in the chemical sciences part II: Outlook
• A Review of Some Techniques for Inclusion of Domain-Knowledge into Deep Neural Networks
• Incorporating Domain Knowledge into Deep Neural Networks
• Deep Learning in Deterministic Computational Mechanics
• Enhancing Discrete Choice Models with Representation Learning
• Characterizing possible failure modes in physics-informed neural networks
• Reinforcement-learning-based control of confined cylinder wakes with stability analyses
• Learning in Sinusoidal Spaces with Physics-Informed Neural Networks
• Is Neuro-Symbolic AI Meeting its Promise in Natural Language Processing? A Structured Review
• Injecting Knowledge in Data-driven Vehicle Trajectory Predictors
• Modulation spectral features for speech emotion recognition using deep neural networks
• Latent Space Data Assimilation by using Deep Learning
• HILL: A Hallucination Identifier for Large Language Models
• An Adaptive Deep Learning Framework for Day-ahead Forecasting of Photovoltaic Power Generation
• Deep Learning with Logical Constraints
• Innovations in Integrating Machine Learning and Agent-Based Modeling of Biomedical Systems
• AI in Human-computer Gaming: Techniques, Challenges and Opportunities
• BubbleNet: Inferring micro-bubble dynamics with semi-physics-informed deep learning
• Explainable Machine Learning with Prior Knowledge: An Overview
• Symbolic Knowledge Extraction and Injection with Sub-symbolic Predictors: A Systematic Literature Review
• Sparse dynamical system identification with simultaneous structural parameters and initial condition estimation
• A machine learning route between band mapping and band structure
• Channel Reflection: Knowledge-Driven Data Augmentation for EEG-Based Brain-Computer Interfaces
• Improving Deep Learning Models via Constraint-Based Domain Knowledge: a Brief Survey
• Quantum Neural Networks for Power Flow Analysis
• Physics-Enhanced Machine Learning: a position paper for dynamical systems investigations
• Toward Physically Plausible Data-Driven Models: A Novel Neural Network Approach to Symbolic Regression
• KID-PPG: Knowledge Informed Deep Learning for Extracting Heart Rate from a Smartwatch
• Curriculum learning for data-driven modeling of dynamical systems
• Physics-Informed Neural Networks for Satellite State Estimation
• Knowledge-integrated AutoEncoder Model
• KnAC: an approach for enhancing cluster analysis with background knowledge and explanations
• Navigating Alignment for Non-identical Client Class Sets: A Label Name-Anchored Federated Learning Framework
• Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling
• Evolutionary Optimization of Physics-Informed Neural Networks: Evo-PINN Frontiers and Opportunities
• Spatial-Temporal Data Mining for Ocean Science: Data, Methodologies, and Opportunities
• Review of Physics-based and Data-driven Multiscale Simulation Methods for Computational Fluid Dynamics and Nuclear Thermal Hydraulics
• Injective Domain Knowledge in Neural Networks for Transprecision Computing
• On the convergence of PINNs
• CorrLoss: Integrating Co-Occurrence Domain Knowledge for Affect Recognition
• Towards Sensor Data Abstraction of Autonomous Vehicle Perception Systems
• KnowTD-An Actionable Knowledge Representation System for Thermodynamics
• Neural Logic Rule Layers
• Quantification of Actual Road User Behavior on the Basis of Given Traffic Rules
• Calibrated simplex-mapping classification
• Guiding Video Prediction with Explicit Procedural Knowledge
• A Quantitative Perspective on Values of Domain Knowledge for Machine Learning
• SUSTeR: Sparse Unstructured Spatio Temporal Reconstruction on Traffic Prediction
• Informed Machine Learning, Centrality, CNN, Relevant Document Detection, Repatriation of Indigenous Human Remains
• A discrete physics-informed training for projection-based reduced order models with neural networks
• Reconstruction of Pairwise Interactions using Energy-Based Models
• Compensating data shortages in manufacturing with monotonicity knowledge
• Non-intrusive model combination for learning dynamical systems
• A Semi-Automated Solution Approach Recommender for a Given Use Case: a Case Study for AI/ML in Oncology via Scopus and OpenAI
• A Clustering-aided Ensemble Method for Predicting Ridesourcing Demand in Chicago
• Informed Machine Learning for Improved Similarity Assessment in Process-Oriented Case-Based Reasoning
• CupNet -- Pruning a network for geometric data
• Physics-Informed Machine Learning in Prognostics and Health Management: A Systematic Literature Review
• Domain Knowledge Based Temporal-Spatial Graph Convolution Network for ECG Recognition
• Neural Ordinary Differential Equations for Learning and Extrapolating System Dynamics Across Bifurcations
• Self-explaining variational posterior distributions for Gaussian Process models