.png){kind=icon}
Return
Modeling polyethylene glycol density using robust soft computing methods
.png){kind=icon}
PRE
AI
DOI：10.1016/j.microc.2025.112815.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Polyethylene Glycol (PEG) is gaining attention as a sustainable green solvent choice for chemical processes and separations, which underpins the need for tools to easily and precisely predict their physio-chemical properties. In this study, eight machine learning methods of Ensemble Learning, Adaptive Boosting, Decision Tree (DT), KNearest Neighbors (KNN), Random Forest (RF), Support Vector Machine (SVM), Multi-Layer Perceptron Artificial Neural Network (MLP-ANN) and Convolutional Neural Network (CNN) are utilized to develop intelligent models to accurately predict the density of PEG in terms of temperature, pressure, and PEG molecular weight gathered from experimental research works published in the literature. An outlier detection algorithm is used to characterize the databank acquired and the relevancy factor is unitized to perform the sensitivity analysis of the input factors on the PEG density. The results show that KNN and RF are identified as the most accurate methods with coefficient of determination equal to 0.986 and 0.984, respectively. The sensitivity analysis indicates that temperature is the most effective input parameter influencing the PEG density. The developed intelligent tools may be used to predict PEG density accurately without requiring arduous and time-consuming experimental tasks.
Keywords:
PEG density
Machine learning
Intelligent models
Relevancy factor
Journal
IF:
5.1
Papers:
1.8W
Citations:
3.7W
Organization
Cited Papers
ShareSaveMachinery Utilization and Production Cost of Wetland, Direct Seeding Paddy Cultivation in Malaysia
OAAI
OAAIShareSaveA model combining a Bayesian network with a modified genetic algorithm for green supplier selection.png)
SIMULATION
SIMULATIONIF0
PREAI
ShareSaveAn Adaptive Machine Learning-based Threat Detection Framework for Industrial Communication Networks
PREAI
ShareSaveShareSaveMolecular insights into the interactions between PEG carriers and drug molecules from Celastrus hindsii: a multi-scale simulation studySCIENTIFIC REPORTSOAAI
SCIENTIFIC REPORTSIF3.9
OAAIShareSave
View more