2019

Explainable Machine Learning for Scientific Insights and Discoveries

Roscher, Ribana, Bohn, Bastian, Duarte, Marco F. et al.

Understand

Machine learning methods have been remarkably successful for a wide range of application areas in the extraction of essential information from data.

An exciting and relatively recent development is the uptake of machine learning in the natural sciences, where the major goal is to obtain novel scientific insights and discoveries from observational or simulated data.
A prerequisite for obtaining a scientific outcome is domain knowledge, which is needed to gain explainability, but also to enhance scientific consistency.
In this article we review explainable machine learning in view of applications in the natural sciences and discuss three core elements which we identified as relevant in this context: transparency, interpretability, and explainability.

Built on

Conditioning by adaptive sampling for robust design
Original
David Brookes, Hahnbeom Park, and Jennifer Listgarten · 1901
Earlier work this paper cites.
Definitions, methods, and applications in interpretable machine learning
Original
W. James Murdoch, Chandan Singh, Karl Kumbier, Reza Abbasi-Asl, and Bin Yu · 1901
Earlier work this paper cites.
Intuitive physics
Michael McCloskey · 1983
Earlier work this paper cites.
Explanation, imagination, and confidence in judgment
Derek J Koehler · 1991
Earlier work this paper cites.
Self-regulation of cognitive inference and decision processes
Roy F. Baumeister and Leonard S. Newman · 1994
Earlier work this paper cites.
Survey and critique of techniques for extracting rules from trained artificial neural networks
Robert Andrews, Joachim Diederich, and Alan B. Tickle · 1995
Earlier work this paper cites.
Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models
Andrea Saltelli, Stefano Tarantola, Francesca Campolongo, and Marco Ratto · 2004
Earlier work this paper cites.
Gaussian Processes for Machine Learning
Carl Edward Rasmussen and Christopher K. I. Williams · 2006
Earlier work this paper cites.
Nonlinear Dimensionality Reduction
John A. Lee and Michel Verleysen · 2007
Earlier work this paper cites.
Kernel methods in machine learning
Thomas Hofmann, Bernhard Schölkopf, and Alexander J. Smola · 2008
Earlier work this paper cites.
Nonnegative Matrix and Tensor Factorization
Andrzej Cichocki, Rafal Zdunek, Anh Huy Phan, and Shun Ichi Amari · 2009
Earlier work this paper cites.
Applications of tensor (multiway array) factorizations and decompositions in data mining
Morten Mørup · 2011
Earlier work this paper cites.
Causality: Models, Reasoning, and Inference, Second Edition
Judea Pearl · 2011
Earlier work this paper cites.
”Explain” in scientific discourse
James A. Overton · 2013
Earlier work this paper cites.
Sparse dynamics for partial differential equations
Hayden Schaeffer, Russel Caflisch, Cory D. Hauck, and Stanley Osher · 2013
Earlier work this paper cites.
Data-driven fluid simulations using regression forests
SoHyeon Jeong, Barbara Solenthaler, Marc Pollefeys, Markus Gross, et al · 2015
Earlier work this paper cites.
Understanding tools: Task-oriented object modeling, learning and recognition
Yixin Zhu, Yibiao Zhao, and Song Chun Zhu · 2015
Earlier work this paper cites.
Discovering governing equations from data by sparse identification of nonlinear dynamical systems
Steven L. Brunton, Joshua L. Proctor, and J. Nathan Kutz · 2016
Earlier work this paper cites.
Retain: An interpretable predictive model for healthcare using reverse time attention mechanism
Edward Choi, Mohammad Taha Bahadori, Jimeng Sun, Joshua Kulas, Andy Schuetz, and Walter Stewart · 2016
Earlier work this paper cites.
Genetic architect: Discovering genomic structure with learned neural architectures
Original
Laura Deming, Sasha Targ, Nate Sauder, Diogo Almeida, and Chun Jimmie Ye · 2016
Earlier work this paper cites.
Feature importance in nonlinear embeddings (fine): applications in digital pathology
Shoshana B Ginsburg, George Lee, Sahirzeeshan Ali, and Anant Madabhushi · 2016
Earlier work this paper cites.
Deep Learning
Ian Goodfellow, Yoshua Bengio, and Aaron Courville · 2016
Earlier work this paper cites.
Learning physical intuition of block towers by example
Adam Lerer, Sam Gross, and Rob Fergus · 2016
Earlier work this paper cites.
To fall or not to fall: A visual approach to physical stability prediction
Original
Wenbin Li, Seyedmajid Azimi, Aleš Leonardis, and Mario Fritz · 2016
Earlier work this paper cites.
Machine learning strategies for systems with invariance properties
Julia Ling, Reese Jones, and Jeremy Templeton · 2016
Earlier work this paper cites.
Reynolds averaged turbulence modelling using deep neural networks with embedded invariance
Julia Ling, Andrew Kurzawski, and Jeremy Templeton · 2016
Earlier work this paper cites.
Inferring biological networks by sparse identification of nonlinear dynamics
N. M. Mangan, S. L. Brunton, J. L. Proctor, and J. N. Kutz · 2016
Earlier work this paper cites.
Accelerating the design of functional glasses through modeling
John C. Mauro, Adama Tandia, K. Deenamma Vargheese, Yihong Z. Mauro, and Morten M. Smedskjaer · 2016
Earlier work this paper cites.
SMASH: Physics-guided reconstruction of collisions from videos
Aron Monszpart, Nils Thuerey, and Niloy J. Mitra · 2016
Earlier work this paper cites.
Newtonian scene understanding: Unfolding the dynamics of objects in static images
Roozbeh Mottaghi, Hessam Bagherinezhad, Mohammad Rastegari, and Ali Farhadi · 2016
Earlier work this paper cites.
A Bayesian approach to calibrating high-throughput virtual screening results and application to organic photovoltaic materials
Edward O. Pyzer-Knapp, Gregor N. Simm, and Alán Aspuru Guzik · 2016
Earlier work this paper cites.
”Why Should I Trust You?”
Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin · 2016
Earlier work this paper cites.
Revealing the vectors of cellular identity with single-cell genomics
Allon Wagner, Aviv Regev, and Nir Yosef · 2016
Earlier work this paper cites.
Discovering phase transitions with unsupervised learning
Lei Wang · 2016
Earlier work this paper cites.
Physics 101: Learning physical object properties from unlabeled videos
J. Wu, J. J. Lim, H. Zhang, J. B. Tenenbaum, and W. T. Freeman · 2016
Earlier work this paper cites.
A compositional object-based approach to learning physical dynamics
Michael B Chang, Tomer Ullman, Antonio Torralba, and Joshua B Tenenbaum · 2017
Earlier work this paper cites.
Towards a rigorous science of interpretable machine learning
Original
Finale Doshi-Velez and Been Kim · 2017
Earlier work this paper cites.
Learning physical descriptors for materials science by compressed sensing
L. M. Ghiringhelli, J. Vybiral, E. Ahmetcik, R. Ouyang, S. V. Levchenko, C. Draxl, and M. Scheffler · 2017
Earlier work this paper cites.
European union regulations on algorithmic decision-making and a “right to explanation”
Bryce Goodman and Seth Flaxman · 2017
Earlier work this paper cites.

Similar

Tropical land use land cover mapping in pará (brazil) using discriminative markov random fields and multi-temporal terrasar-x data
Ron Hagensieker, Ribana Roscher, Johannes Rosentreter, Benjamin Jakimow, and Björn Waske · 2017
Cited alongside, same era.
An information-based machine learning approach to elasticity imaging
Cameron Hoerig, Jamshid Ghaboussi, and Michael F. Insana · 2017
Cited alongside, same era.
Discovering phases, phase transitions, and crossovers through unsupervised machine learning: A critical examination
Wenjian Hu, Rajiv RP Singh, and Richard T Scalettar · 2017
Cited alongside, same era.
Theory-guided data science: A new paradigm for scientific discovery from data
A. Karpatne, G. Atluri, J. H. Faghmous, M. Steinbach, A. Banerjee, A. Ganguly, S. Shekhar, N. Samatova, and V. Kumar · 2017
Cited alongside, same era.
Object-oriented deep learning
Qianli Liao and Tomaso Poggio · 2017
Cited alongside, same era.
Design principles and top non-fullerene acceptor candidates for organic photovoltaics
Steven A. Lopez, Benjamin Sanchez-Lengeling, Julio de Goes Soares, and Alán Aspuru-Guzik · 2017
Cited alongside, same era.
Methods for interpreting and understanding deep neural networks
Grégoire Montavon, Wojciech Samek, and Klaus-Robert Müller · 2017
Cited alongside, same era.
Machine learning of linear differential equations using Gaussian processes
Maziar Raissi, Paris Perdikaris, and George Em Karniadakis · 2017
Cited alongside, same era.
Data-driven discovery of partial differential equations
Samuel H. Rudy, Steven L. Brunton, Joshua L. Proctor, and J. Nathan Kutz · 2017
Cited alongside, same era.
Quantum-chemical insights from deep tensor neural networks
Kristof T Schütt, Farhad Arbabzadah, Stefan Chmiela, Klaus R Müller, and Alexandre Tkatchenko · 2017
Cited alongside, same era.
Attend and predict: Understanding gene regulation by selective attention on chromatin
Ritambhara Singh, Jack Lanchantin, Arshdeep Sekhon, and Yanjun Qi · 2017
Cited alongside, same era.
Label-free supervision of neural networks with physics and domain knowledge
R. Stewart and S. Ermon · 2017
Cited alongside, same era.

Then

Deep learning for universal linear embeddings of nonlinear dynamics
Bethany Lusch, J. Nathan Kutz, and Steven L. Brunton · 2018
Later among the works it cites.
Using deep learning to model the hierarchical structure and function of a cell
Jianzhu Ma, Michael Ku Yu, Samson Fong, Keiichiro Ono, Eric Sage, Barry Demchak, Roded Sharan, and Trey Ideker · 2018
Later among the works it cites.
A regression approach for explaining manifold embedding coordinates
Original
Marina Meila, Samson Koelle, and Hanyu Zhang · 2018
Later among the works it cites.
Explanation in artificial intelligence: Insights from the social sciences
Tim Miller · 2018
Later among the works it cites.
The building blocks of interpretability
Chris Olah, Arvind Satyanarayan, Ian Johnson, Shan Carter, Ludwig Schubert, Katherine Ye, and Alexander Mordvintsev · 2018
Later among the works it cites.
SISSO: A compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates
Runhai Ouyang, Stefano Curtarolo, Emre Ahmetcik, Matthias Scheffler, and Luca M. Ghiringhelli · 2018
Later among the works it cites.
Individual brain charting, a high-resolution fmri dataset for cognitive mapping
Ana Luísa Pinho, Alexis Amadon, Torsten Ruest, Murielle Fabre, Elvis Dohmatob, Isabelle Denghien, Chantal Ginisty, Séverine Becuwe-Desmidt, Séverine Roger, Laurence Laurier, Véronique Joly-Testault, Gaëlle Médiouni-Cloarec, Christine Doublé, Bernadette Martins, Philippe Pinel, Evelyn Eger, G. Varoquaux, Christophe Pallier, Stanislas Dehaene, Lucie Hertz-Pannier, and Bertrand Thirion · 2018
Later among the works it cites.
Deep hidden physics models: Deep learning of nonlinear partial differential equations
Maziar Raissi · 2018
Later among the works it cites.
Explainable artificial intelligence: Understanding, visualizing and interpreting deep learning models
Wojciech Samek, Thomas Wiegand, and Klaus-Robert Müller · 2018
Later among the works it cites.
Detecting statistical interactions from neural network weights
Michael Tsang, Dehua Cheng, and Yan Liu · 2018
Later among the works it cites.
Atlases of cognition with large-scale human brain mapping
Gaël Varoquaux, Yannick Schwartz, Russell A. Poldrack, Baptiste Gauthier, Danilo Bzdok, Jean-Baptiste Poline, and Bertrand Thirion · 2018
Later among the works it cites.
The challenge of crafting intelligible intelligence
D Weld and Gagan Bansal · 2018
Later among the works it cites.
Interpretable intuitive physics model
Tian Ye, Xiaolong Wang, James Davidson, and Abhinav Gupta · 2018
Later among the works it cites.
Data-Driven Science and Engineering
Steven L. Brunton and J. Nathan Kutz · 2019
Closest in time.
Machine learning and the physical sciences
Giuseppe Carleo, Ignacio Cirac, Kyle Cranmer, Laurent Daudet, Maria Schuld, Naftali Tishby, Leslie Vogt-Maranto, and Lenka Zdeborová · 2019
Closest in time.
Interpretable machine learning for inferring the phase boundaries in a nonequilibrium system
C. Casert, T. Vieijra, J. Nys, and J. Ryckebusch · 2019
Closest in time.
Automated, predictive, and interpretable inference of caenorhabditis elegans escape dynamics
Bryan C Daniels, William S Ryu, and Ilya Nemenman · 2019
Closest in time.
Hyperspectral plant disease forecasting using generative adversarial networks
Alina Förster, Jens Behley, Jan Behmann, and Ribana Roscher · 2019
Closest in time.
From genotype to phenotype: Augmenting deep learning with networks and systems biology
Vahid H. Gazestani and Nathan E. Lewis · 2019
Closest in time.
Probing hidden spin order with interpretable machine learning
Jonas Greitemann, Ke Liu, and Lode Pollet · 2019
Closest in time.
Manifold learning for parameter reduction
A. Holiday, M. Kooshkbaghib, J. M. Bello-Rivas, C. W. Gear, A. Zagaris, and I. G. Kevrekidis · 2019
Closest in time.
Causability and explainabilty of artificial intelligence in medicine
Andreas Holzinger, Georg Langs, Helmut Denk, Kurt Zatloukal, and Heimo Müller · 2019
Closest in time.
Reliable and explainable machine learning methods for accelerated material discovery
Bhavya Kailkhura, Brian Gallagher, Sookyung Kim, Anna Hiszpanski, and T Han · 2019
Closest in time.
Unmasking Clever Hans predictors and assessing what machines really learn
Sebastian Lapuschkin, Stephan Wäldchen, Alexander Binder, Grégoire Montavon, Wojciech Samek, and Klaus-Robert Müller · 2019
Closest in time.
Neural-guided symbolic regression with semantic prior
Original
Li Li, Minjie Fan, Rishabh Singh, and Patrick Riley · 2019
Closest in time.
Learning multiple order parameters with interpretable machines
Ke Liu, Jonas Greitemann, and Lode Pollet · 2019
Closest in time.
A rate-distortion framework for explaining neural network decisions
Original
Jan MacDonald, Stephan Wäldchen, Sascha Hauch, and Gitta Kutyniok · 2019
Closest in time.
Explaining explanations in AI
Brent Mittelstadt, Chris Russell, and Sandra Wachter · 2019
Closest in time.
Deep learning and process understanding for data-driven Earth system science
Markus Reichstein, Gustau Camps-Valls, Bjorn Stevens, Martin Jung, Joachim Denzler, Nuno Carvalhais, et al · 2019
Closest in time.
Interpretations are useful: penalizing explanations to align neural networks with prior knowledge
Original
Laura Rieger, Chandan Singh, W James Murdoch, and Bin Yu · 2019
Closest in time.
Attention gated networks: Learning to leverage salient regions in medical images
Jo Schlemper, Ozan Oktay, Michiel Schaap, Mattias Heinrich, Bernhard Kainz, Ben Glocker, and Daniel Rueckert · 2019
Closest in time.
Rethinking drug design in the artificial intelligence era
Petra Schneider, W. Patrick Walters, Alleyn T. Plowright, Norman Sieroka, Jennifer Listgarten, Robert A. Goodnow, Jasmin Fisher, Johanna M. Jansen, José S. Duca, Thomas S. Rush, Matthias Zentgraf, John Edward Hill, Elizabeth Krutoholow, Matthias Kohler, Jeff Blaney, Kimito Funatsu, Chris Luebkemann, and Gisbert Schneider · 2019
Closest in time.
Causality for machine learning
Original
Bernhard Schölkopf · 2019
Closest in time.
From deep learning to mechanistic understanding in neuroscience: the structure of retinal prediction
Hidenori Tanaka, Aran Nayebi, Niru Maheswaranathan, Lane McIntosh, Stephen Baccus, and Surya Ganguli · 2019
Closest in time.
Attention-based deep neural networks for detection of cancerous and precancerous esophagus tissue on histopathological slides
Naofumi Tomita, Behnaz Abdollahi, Jason Wei, Bing Ren, Arief Suriawinata, and Saeed Hassanpour · 2019
Closest in time.
Informed machine learning – towards a taxonomy of explicit integration of knowledge into machine learning
Original
Laura von Rueden, Sebastian Mayer, Jochen Garcke, Christian Bauckhage, and Jannis Schuecker · 2019
Closest in time.
Prediction of Reynolds stresses in high-Mach-number turbulent boundary layers using physics-informed machine learning
Jian Xun Wang, Junji Huang, Lian Duan, and Heng Xiao · 2019
Closest in time.
GroupINN: Grouping-based interpretable neural network for classification of limited, noisy brain data
Yujun Yan, Jiong Zhu, Marlena Duda, Eric Solarz, Chandra Sripada, and Danai Koutra · 2019
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Reading the bibliography…