Feature Selection for Polymer Informatics: Evaluating Scalability and Robustness of the FS4RV<sub>DD</sub> Algorithm Using Synthetic Polydisperse Data Sets

Cravero, Fiorella; Schustik, Santiago; Martínez, María Jimena; Vazquez, Gustavo E.; Díaz, Mónica F.; Ponzoni, Ignacio

doi:10.1021/acs.jcim.9b00867

Cited by 11 publications

(7 citation statements)

References 50 publications

(90 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After comparing 79 model setups using 3 different polymer representations, 7 feature representations, and 8 ML algorithms, we reveal the pros and cons behind different options. Such a systematic benchmark study provides valuable guidance for the model selection in polymer’s T g prediction, and it would also benefit other polymer informatics tasks such as mechanical, electronic, or optical properties. − …”

Section: Introductionmentioning

confidence: 99%

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature

Tao

Varshney

2021

J. Chem. Inf. Model.

128

View full text Add to dashboard Cite

In the field of polymer informatics, utilizing machine learning (ML) techniques to evaluate the glass transition temperature T g and other properties of polymers has attracted extensive attention. This data-centric approach is much more efficient and practical than the laborious experimental measurements when encountered a daunting number of polymer structures. Various ML models are demonstrated to perform well for T g prediction. Nevertheless, they are trained on different data sets, using different structure representations, and based on different feature engineering methods. Thus, the critical question arises on selecting a proper ML model to better handle the T g prediction with generalization ability. To provide a fair comparison of different ML techniques and examine the key factors that affect the model performance, we carry out a systematic benchmark study by compiling 79 different ML models and training them on a large and diverse data set. The three major components in setting up an ML model are structure representations, feature representations, and ML algorithms. In terms of polymer structure representation, we consider the polymer monomer, repeat unit, and oligomer with longer chain structure. Based on that feature, representation is calculated, including Morgan fingerprinting with or without substructure frequency, RDKit descriptors, molecular embedding, molecular graph, etc. Afterward, the obtained feature input is trained using different ML algorithms, such as deep neural networks, convolutional neural networks, random forest, support vector machine, LASSO regression, and Gaussian process regression. We evaluate the performance of these ML models using a holdout test set and an extra unlabeled data set from high-throughput molecular dynamics simulation. The ML model’s generalization ability on an unlabeled data set is especially focused, and the model’s sensitivity to topology and the molecular weight of polymers is also taken into consideration. This benchmark study provides not only a guideline for the T g prediction task but also a useful reference for other polymer informatics tasks.

show abstract

Section: Introductionmentioning

confidence: 99%

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature

Tao

Varshney

2021

J. Chem. Inf. Model.

128

View full text Add to dashboard Cite

show abstract

“…Several researchers focused on developing a method for analyzing the performance of several polymers including CPE. That are all can be depicted below, Cravero et al (2019) [8] examine the FS4RVDD algorithm's scalability and reliability. The measurement of the data bank, the modulo of the aspect subsets, noise in data, and kind of correlation were all varied and combined to create synthetic data (linear and nonlinear).…”

Section: Related Workmentioning

confidence: 99%

Evaluation of Performance of Chlorinated Polyethylene Using Wireless Network and Artificial Intelligence Technology

Zhang

Zhou

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

Chemical enterprises are presently confronted with several difficult issues, including high power consumption, dangerous risk evaluation, and environmental regulation, all of which push industrial and academic institutions to develop new technologies, catalysts, and materials. Chlorinated polyethylene (CPE) is a polymer made by replacing H2 molecules in high density-(C2H4)n with chloride ions. CPE elastomers are made from a high density-(C2H4) backbone, and it was chlorinated using a free radical aqueous slurry technique. However, such fundamental polymer characteristics are insufficient to explain the performance characteristics of chlorinated polyethylene elastomers. Artificial intelligence (AI) has had a massive effect on all sections of the chemical sector, with tremendous potential that has revolutionized value supply chains, enhanced efficiency, and opened up new ways to the marketplace. As a result, in this research, we offer a methodology for the performance characterization of chlorinated polyethylene based on artificial intelligence (AI) and wireless network technology. The AI tools can search through enormous databases of known compounds and their attributes, leveraging the data to generate new possibilities. The dataset is first gathered. The chemical characterization is classified using the K -nearest neighbor (KNN) technique. This program was created to examine molecule structures and forecast the outcomes of new chemical reactions. Bayesian optimization is used to improve characterization performance. The proposed method will contribute to the future usage of AI in the chemical sector.

show abstract

“…Many groups have conducted experiments and used polymer informatics to research polymers actively and to enhance their performance or discover new polymers. Recently, machine learning has been applied to polymer informatics, and this has produced remarkable results [1][2][3][4][5]. Thus, the progress of machine learning algorithms is important for polymer informatics; however, the availability of systematic datasets for machine learning is more important [6].…”

Section: Introductionmentioning

confidence: 99%

Machine extraction of polymer data from tables using XML versions of scientific articles

Oka

Yoshizawa

Shindo

et al. 2021

Science and Technology of Advanced Materials: Methods

View full text Add to dashboard Cite

In this study, we examined machine extraction of polymer data from tables in scientific articles. The extraction system consists of five processes: table extraction, data formatting, polymer name recognition, property specifier identification, and data extraction. Tables were first extracted in plain text. XML versions of scientific articles were used, and the tabular forms were accurately extracted, even for complicated tables, such as multi-column, multi-row, and merged tables. Polymer name recognition was performed using a named entity recognizer created by deep neural network learning of polymer names. The preparation cost of the training data was reduced using a rule-based algorithm. The target polymer properties in this study were glass transition temperature (T g ), melting temperature (T m ), and decomposition temperature (T d ), and the specifiers were identified using partial string matching. Through these five processes, 2,181 data points for T g , 1,526 for T m , and 2,316 for T d were extracted from approximately 18,000 scientific articles published by Elsevier. Nearly half of them were extracted from complicated tables. The F-scores for the extraction were 0.871, 0.870, and 0.841, respectively. These results indicate that the extraction system created in this study can rapidly and accurately collect large amounts of polymer data from tables in scientific literature.

show abstract

Feature Selection for Polymer Informatics: Evaluating Scalability and Robustness of the FS4RV_DD Algorithm Using Synthetic Polydisperse Data Sets

Cited by 11 publications

References 50 publications

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature

Evaluation of Performance of Chlorinated Polyethylene Using Wireless Network and Artificial Intelligence Technology

Machine extraction of polymer data from tables using XML versions of scientific articles

Contact Info

Product

Resources

About

Feature Selection for Polymer Informatics: Evaluating Scalability and Robustness of the FS4RVDD Algorithm Using Synthetic Polydisperse Data Sets

Cited by 11 publications

References 50 publications

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature

Benchmarking Machine Learning Models for Polymer Informatics: An Example of Glass Transition Temperature

Evaluation of Performance of Chlorinated Polyethylene Using Wireless Network and Artificial Intelligence Technology

Machine extraction of polymer data from tables using XML versions of scientific articles

Contact Info

Product

Resources

About

Feature Selection for Polymer Informatics: Evaluating Scalability and Robustness of the FS4RV_DD Algorithm Using Synthetic Polydisperse Data Sets