Machine learning has been increasingly used in biochemistry.However,in organic chemistry and other experiment-based fields,data collected from real experiments are inadequate and the current coronavirus disease(COVID-...Machine learning has been increasingly used in biochemistry.However,in organic chemistry and other experiment-based fields,data collected from real experiments are inadequate and the current coronavirus disease(COVID-19)pandemic has made the situation even worse.Such limited data resources may result in the low performance of modeling and affect the proper development of a control strategy.This paper proposes a feasible machine learning solution to the problem of small sample size in the biopolymerization process.To avoid overfitting,the variational auto-encoder and generative adversarial network algorithms are used for data augmentation.The random forest and artificial neural network algorithms are implemented in the modeling process.The results prove that data augmentation techniques effectively improve the performance of the regression model.Several machine learning models were compared and the experimental results show that the random forest model with data augmentation by the generative adversarial network technique achieved the best performance in predicting the molecular weight on the training set(with an R^(2) of 0.94)and on the test set(with an R^(2) of 0.74),and the coefficient of determination of this model was 0.74.展开更多
Due to their lower environmental impact, ease of accessibility, low cost, and biodegradability, bio-renewable sources have been used extensively in the last several decades to synthesize adhesives, substituting petroc...Due to their lower environmental impact, ease of accessibility, low cost, and biodegradability, bio-renewable sources have been used extensively in the last several decades to synthesize adhesives, substituting petrochemical-based adhesive. Vegetable oils (including palm, castor, jatropha, and soybean oils), lactic acid, potato starch, and other bio-renewable sources are all excellent sources for the synthesis of adhesives that are being taken into consideration for the synthesis of “eco-friendly” adhesives. Due to their widespread use, accessibility, affordability, and biodegradability, biobased raw materials like carbohydrates used to synthesize wood and wood composite adhesive have gradually replaced petrochemical-based adhesive. Recently, xanthan gum, a naturally occurring polymer, has drawn the interest of scientists as a potentially petroleum source replacement. It possesses specific rheological characteristics, excellent water solubility, and stability to heat, and can be used as a binder, thickener, suspending agent, and stabilizer. Xanthan gum increases the adhesive strength in addition to increasing the viscosity of water-soluble adhesives. This article discusses xanthan gum as a potential substitute for traditional raw materials derived from petroleum that is used as a raw material for adhesives.展开更多
文摘Machine learning has been increasingly used in biochemistry.However,in organic chemistry and other experiment-based fields,data collected from real experiments are inadequate and the current coronavirus disease(COVID-19)pandemic has made the situation even worse.Such limited data resources may result in the low performance of modeling and affect the proper development of a control strategy.This paper proposes a feasible machine learning solution to the problem of small sample size in the biopolymerization process.To avoid overfitting,the variational auto-encoder and generative adversarial network algorithms are used for data augmentation.The random forest and artificial neural network algorithms are implemented in the modeling process.The results prove that data augmentation techniques effectively improve the performance of the regression model.Several machine learning models were compared and the experimental results show that the random forest model with data augmentation by the generative adversarial network technique achieved the best performance in predicting the molecular weight on the training set(with an R^(2) of 0.94)and on the test set(with an R^(2) of 0.74),and the coefficient of determination of this model was 0.74.
文摘Due to their lower environmental impact, ease of accessibility, low cost, and biodegradability, bio-renewable sources have been used extensively in the last several decades to synthesize adhesives, substituting petrochemical-based adhesive. Vegetable oils (including palm, castor, jatropha, and soybean oils), lactic acid, potato starch, and other bio-renewable sources are all excellent sources for the synthesis of adhesives that are being taken into consideration for the synthesis of “eco-friendly” adhesives. Due to their widespread use, accessibility, affordability, and biodegradability, biobased raw materials like carbohydrates used to synthesize wood and wood composite adhesive have gradually replaced petrochemical-based adhesive. Recently, xanthan gum, a naturally occurring polymer, has drawn the interest of scientists as a potentially petroleum source replacement. It possesses specific rheological characteristics, excellent water solubility, and stability to heat, and can be used as a binder, thickener, suspending agent, and stabilizer. Xanthan gum increases the adhesive strength in addition to increasing the viscosity of water-soluble adhesives. This article discusses xanthan gum as a potential substitute for traditional raw materials derived from petroleum that is used as a raw material for adhesives.
