Using Shapley additive explanations to interpret extreme gradient boosting predictions of grassland degradation in Xilingol, China by Batunacun, Batunacun, R. Wieland, T. Lakes, T. Lakes, C. Nendel, C. Nendel

“…A set of 20 drivers was analysed using XGBoost, involving four alternative sampling strategies, and SHAP (Shapley additive explanations) to interpret the results of the purely data-driven approach. The results indicated that, with three of the sampling strategies (over-balanced, balanced, and imbalanced), XGBoost achieved similar and robust simulation results. …”
Get full text

Modular organization of functional network connectivity in healthy controls and patients with schizophrenia during the resting state by Qingbao eYu, Sergey M Plis, Erik B Erhardt, Elena A Allen, Jing eSui, Kent A Kiehl, Kent A Kiehl, Godfrey ePearlson, Godfrey ePearlson, Vince D Calhoun, Vince D Calhoun, Vince D Calhoun

“…However, the organization of functional network connectivity (FNC), comprising a purely data-driven network built from spatially independent brain components, is not yet clear. …”
Get full text

A review of mechanistic learning in mathematical oncology by John Metzcar, John Metzcar, Catherine R. Jutzeler, Catherine R. Jutzeler, Paul Macklin, Alvaro Köhn-Luque, Alvaro Köhn-Luque, Sarah C. Brüningk, Sarah C. Brüningk

“…We highlight the synergistic potential of mechanistic learning and point out similarities and differences between purely data-driven and mechanistic approaches concerning model complexity, data requirements, outputs generated, and interpretability of the algorithms and their results. …”
Get full text

A machine learning framework to quantify and assess the impact of COVID-19 on the power sector: An Indian context by Manu Suvarna, Apoorva Katragadda, Ziying Sun, Yun Bin Choh, Qianyu Chen, Pravin PS, Xiaonan Wang

“…The modeling framework developed in this study is purely data-orientated, cross-deployable across spatio-temporal scales and can serve as a valuable tool to inform current and future energy policies amidst and post COVID-19.…”
Get full text

Forecasting of Mesoscale Eddies in the Kuroshio Extension Based on Temporal Modes-Enhanced Neural Network by Haitong Wang, Yunxia Guo, Yuan Kong, Yong Fang

“…Mesoscale eddies are a common occurrence in the Kuroshio Extension (KE) that have a major impact on the levels of salinity and heat transport in the Northwest Pacific, the strength of the Kuroshio jet, and the fluctuations of the Kuroshio’s trajectory. In this study, a purely data-driven machine learning model, Temporal Modes-Enhanced Neural Network (TMENN), is proposed to forecast the spatiotemporal variation of mesoscale eddies based on daily sea surface height (SSH) data over a 20-year period (2000–2019) in the Kuroshio Extension. …”
Get full text

An integrative gene selection with association analysis for microarray data classification by Ong, Huey Fang, Mustapha, Norwati, Sulaiman, Md. Nasir

“…The rising interest in integrative approach has shifted gene selection from purely data-centric to incorporating additional biological knowledge. …”
Get full text

A Deep Learning Algorithm for the Max-Cut Problem Based on Pointer Network Structure with Supervised Learning and Reinforcement Learning Strategies by Shenshen Gu, Yue Yang

“…A pointer network is a sequence-to-sequence deep neural network, which can extract data features in a purely data-driven way to discover the hidden laws behind data. …”
Get full text

Using machine learning to model uncertainty for water vapor atmospheric motion vectors by J. V. Teixeira, H. Nguyen, D. J. Posselt, H. Su, L. Wu

“…The method presented in this paper supplements existing approaches to error specification by providing an error characterization module that is purely data-driven. Our proposed error characterization method combines the flexibility of machine learning (random forest) with the robust error estimates of unsupervised parametric clustering (using a Gaussian mixture model). …”
Get full text

Modeling human cancer-related regulatory modules by GA-RNN hybrid algorithms by Chao Shih-Yi, Chiang Jung-Hsien

“…In addition, this inverse modeling correctly identifies known oncogenes and their interaction genes in a purely data-driven way.</p>…”
Get full text

Physics-informed neural ODE (PINODE): embedding physics into models using collocation points by Aleksei Sholokhov, Yuying Liu, Hassan Mansour, Saleh Nabi

“…Namely, for a problem of modeling a high-dimensional nonlinear PDE, our experiments show $$\times$$ × 5 performance gains, measured by prediction error, in a low-data regime, $$\times$$ × 10 performance gains in tasks of high-noise learning, $$\times$$ × 100 gains in the efficiency of utilizing the latent-space dimension, and $$\times$$ × 200 gains in tasks of far-out out-of-distribution forecasting relative to purely data-driven models. These improvements pave the way for broader adoption of network-based physics-informed ROMs in compressive sensing and control applications.…”
Get full text

Learning Low-Precision Structured Subnetworks Using Joint Layerwise Channel Pruning and Uniform Quantization by Xinyu Zhang, Ian Colbert, Srinjoy Das

“…Informed by this observation, we propose a purely data-driven nonparametric, magnitude-based channel pruning strategy that works in a greedy manner based on the activations of the previous sparsified layer. …”
Get full text

Discovering interpretable physical models using symbolic regression and discrete exterior calculus by Simone Manti, Alessandro Lucantonio

“…While access to large amount of data has fueled the use of machine learning to recover physical models from experiments and increase the accuracy of physical simulations, purely data-driven models have limited generalization and interpretability. …”
Get full text

FaIRGP: A Bayesian Energy Balance Model for Surface Temperatures Emulation by Bouabid, S, Sejdinovic, D, Watson‐Parris, D

“…Its ability to learn from data allows it to outperform EBMs, while its robust physical foundation safeguards against the pitfalls of purely data‐driven models. We also illustrate how FaIRGP can be used to obtain estimates of top‐of‐atmosphere radiative forcing and discuss the benefits of its mathematical tractability for applications such as detection and attribution or precipitation emulation. …”

Learning pharmacokinetic models for in vivo glucocorticoid activation by Bunte, K, Smith, D, Chappell, M, Hassan-Smith, Z, Tomlinson, J, Arlt, W, Tiňo, P

“…To understand trends in individual responses to medication, one can take a purely data-driven machine learning approach, or alternatively apply pharmacokinetics combined with mixed-effects statistical modelling. …”

Uncertainty quantified discovery of chemical reaction systems via Bayesian scientific machine learning by Emily Nieves, Emily Nieves, Raj Dandekar, Chris Rackauckas

“…Additionally, we demonstrate how the embedding of scientific knowledge improves extrapolation accuracy by comparing results to purely data-driven machine learning methods. Together, this provides a new framework for robust, autonomous Bayesian inference on unknown or complex chemical and biological reaction systems.…”
Get full text

Regional parent flood frequency distributions in Europe – Part 2: Climate and scale controls by J. L. Salinas, A. Castellarin, S. Kohnová, T. R. Kjeldsen

“…A database of L-moment ratios of annual maximum series (AMS) of peak discharges from Austria, Italy and Slovakia, involving a total of 813 catchments with more than 25 yr of record length is presented, together with mean annual precipitation (MAP) and basin area as catchment descriptors surrogates of climate and scale controls. A purely data-based investigation performed on the database shows that the <i>generalized extreme value</i> (GEV) distribution provides a better representation of the averaged sample L-moment ratios compared to the other distributions considered, for catchments with medium to higher values of MAP independently of catchment area, while the <i>three-parameter lognormal</i> distribution is probably a more appropriate choice for drier (lower MAP) intermediate-sized catchments, which presented higher skewness values. …”
Get full text

Leveraging sanitized data for probabilistic electricity market prediction: a Singapore case study by Ning Zhou Xu, Xiang Gao, Songjian Chai, Ming Niu, Jia Xin Yang

“…These insights are challenging to obtain using purely data-driven methods. This paper proposes a physics-based solution for the probabilistic prediction of market-clearing outcomes, using real sanitized offer data from the National Electricity Market of Singapore (NEMS). …”
Get full text

Temperature Prediction of Mushrooms Based on a Data—Physics Hybrid Approach by Mingfei Wang, Xiangshu Kong, Feifei Shan, Wengang Zheng, Pengfei Ren, Jiaoling Wang, Chunling Chen, Xin Zhang, Chunjiang Zhao

“…The results demonstrated that integrating a simplified physical model into the predictive model based on data decomposition led to a 12.50% reduction in the RMSE of the model’s predictions compared to a purely data-driven model. The model proposed in this article exhibited good predictive performance in small datasets, reducing the time required for data collection in modeling.…”
Get full text

MetaMGC: a music generation framework for concerts in metaverse by Cong Jin, Fengjuan Wu, Jing Wang, Yang Liu, Zixuan Guan, Zhe Han

“…In addition, this paper validates a neural rendering method that can be used to generate spatial audio based on a binaural-integrated neural network with a fully convolutional technique. And the purely data-driven end-to-end model performs to be more reliable compared with traditional spatial audio generation methods such as HRTF. …”
Get full text

Investigation of Feature Engineering Methods for Domain-Knowledge-Assisted Bearing Fault Diagnosis by Christoph Bienefeld, Florian Michael Becker-Dombrowsky, Etnik Shatri, Eckhard Kirchner

“…In current research, purely data-driven deep learning methods are becoming increasingly popular, aiming for accurate predictions of bearing faults without requiring bearing-specific domain knowledge. …”
Get full text