Maximizing information from chemical engineering data sets: Applications to machine learning
arXiv:2201.10035 · doi:10.1016/j.ces.2022.117469
Abstract
It is well-documented how artificial intelligence can have (and already is having) a big impact on chemical engineering. But classical machine learning approaches may be weak for many chemical engineering applications. This review discusses how challenging data characteristics arise in chemical engineering applications. We identify four characteristics of data arising in chemical engineering applications that make applying classical artificial intelligence approaches difficult: (1) high variance, low volume data, (2) low variance, high volume data, (3) noisy/corrupt/missing data, and (4) restricted data with physics-based limitations. For each of these four data characteristics, we discuss applications where these data characteristics arise and show how current chemical engineering research is extending the fields of data science and machine learning to incorporate these challenges. Finally, we identify several challenges for future research.
arXiv comment34 pages, 3 figures, 1 table
References in corpus (20)
- On representing chemical environments
- Big Data of Materials Science - Critical Role of the Descriptor
- GAIN: Missing Data Imputation using Generative Adversarial Nets
- Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations
- Optimization under Uncertainty in the Era of Big Data and Deep Learning: When Machine Learning Meets Mathematical Programming
- Resolving transition metal chemical space: feature selection for machine learning and structure-property relationships
- Understanding molecular representations in machine learning: The role of uniqueness and target similarity
- Computer-aided molecular design: An introduction and review of tools, applications, and solution techniques
- The ALAMO approach to machine learning
- Global Deterministic Optimization with Artificial Neural Networks Embedded
- Learning Sparse Neural Networks through Regularization
- Stochastic data-driven model predictive control using Gaussian processes
- Data-Driven Stochastic Robust Optimization: A General Computational Framework and Algorithm for Optimization under Uncertainty in the Big Data Era
- A Unified View of Piecewise Linear Neural Network Verification
- ENTMOOT: A Framework for Optimization over Ensemble Tree Models
- Integrating production scheduling and process control using latent variable dynamic models
- Distributionally Robust Chance Constrained Programming with Generative Adversarial Networks (GANs)
- Multi-Objective Constrained Optimization for Energy Applications via Tree Ensembles
- Partition-based formulations for mixed-integer optimization of trained ReLU neural networks
- Data-Driven Robust Optimization using Unsupervised Deep Learning