Accelerating the Selection of Covalent Organic Frameworks with Automated Machine Learning

Abstract

Covalent organic frameworks (COFs) have the advantages of high thermal stability and large specific surface and have great application prospects in the fields of gas storage and catalysis. This article mainly focuses on COFs' working capacity of methane (CH 4 ). Due to the vast number of possible COF structures, it is time-consuming to use traditional calculation methods to find suitable materials, so it is important to apply appropriate machine learning (ML) algorithms to build accurate prediction models. A major obstacle for the use of ML algorithms is that the performance of an algorithm may be affected by many design decisions. Finding appropriate algorithm and model parameters is quite a challenge for nonprofessionals. In this work, we use automated machine learning (AutoML) to analyze the working capacity of CH 4 based on 403,959 COFs. We explore the relationship between 23 features such as the structure, chemical characteristics, atom types of COFs, and the working capacity. Then, the tree-based pipeline optimization tool (TPOT) in AutoML and the traditional ML methods including multiple linear regression, support vector machine, decision tree, and random forest that manually set model parameters are compared. It is found that the TPOT can not only save complex data preprocessing and model parameter tuning but also show higher performance than traditional ML models. Compared with traditional grand canonical Monte Carlo simulations, it can save a lot of time. AutoML has broken through the limitations of professionals so that researchers in nonprofessional fields can realize automatic parameter configuration for experiments to obtain highly accurate and easy-to-understand results, which is of great significance for material screening.

Combining Machine Learning and Molecular Simulations to Unlock Gas Separation Potentials of MOF Membranes and MOF/Polymer MMMs

“…The Pearson correlation coefficient ( r ) was used to determine the feature correlations, which can be expressed as , where x and y are the features, and x̅ and y̅ are the means of x and y . If the two descriptors are strongly correlated, it can cause problems such as multicollinearity and overtraining of ML models . To avoid these, we computed the r values between each descriptor and removed the one having a strong correlation ( r > 0.90).…”

“…In TPOT, a random principal singular value decomposition variant called randomized principal component analysis (PCA) is used for feature extraction. Comparison of a CH 4 working capacity of 403,959 hypothetical COFs predicted using the algorithms defined by TPOT and traditional ML models such as decision tree (DT), random forest (RF), and support vector machine (SVM) showed that the accuracy of ML predictions obtained from TPOT is higher than those of traditional ML models . For the model selection in TPOT, the regression algorithms in the scikit-learn toolkit were used.…”

Machine-Learning-Aided Computational Study of Covalent Organic Frameworks for Reversed C2H6/C2H4 Separation

“…Nowadays, machine learning is being used to predict material properties and filter materials with high performances from databases. ,, Meanwhile, adsorption selectivity has been generally considered as the most critical factor to rank porous materials for gas purification. ,, Herein, to discover a better ML-based method suitable for predicting selectivity, especially, to gain a deep insight into the ranking of the influence of various descriptors of COFs on its selectivity, four ML algorithms including DT, RF, SVM, and ANN , were trained and tested. As shown in Figure , the RF algorithm shows the highest R 2 value (0.970) and the lowest MAE (0.674) and RMSE (1.051) values, indicating that the predicted S ads,C 2 H 6 /C 2 H 4 using the RF algorithm are more consistent with the simulated results.…”

Light-Induced Organic Transformations by Covalent Organic Frameworks as Reticular Platforms for Selective Photosynthesis

“…As a matter of fact, the ideal COF-based photocatalyst, tailored for a specific reaction, might be already described in the literature, in terms of its molecular composition and framework arrangement. In this context, the emerging power of active machine learning (AML) approaches − will probably be of great help in assisting and accelerating the discovery of the best suitable COF-based photocatalyst, for instance by analyzing charge transport, optical absorption range, and linkage bond strength, among other specific descriptors. Indeed, fundamental photophysical studies can provide important guidelines for photocatalytic COF design.…”