Despite the usage of both experimental and topological methods, realizing a rapid and accurate measurement of the onset temperature(Tg ) of GexSe1-xglass transition remains an open challenge. In this paper, a predicti...Despite the usage of both experimental and topological methods, realizing a rapid and accurate measurement of the onset temperature(Tg ) of GexSe1-xglass transition remains an open challenge. In this paper, a predictive model for the Tg in GexSe1-xglass system is presented by a machine learning method named feature selection based two-stage support vector regression(FSTS-SVR). Firstly, Pearson correlation coefficient(PCC) is used to select features highly correlated with Tg from the candidate features of GexSe1-x glass system. Secondly, in order to simulate the two-stage characteristic of Tg which is caused by structural variation with a turning point at x = 0.33 via the structural analysis, SVR is utilized to build predictive models for two stages separately and then the two achieved models are synthesized using a minimum error based model for Tg prediction. Compared with the topological and other methods based on SVR, the FSTS-SVR gives the highest predictive accuracy with the root mean square error(RMSE) and mean absolute percentage error(MAPE) of 10.64 K and 2.38%, respectively. This method is also expected to be more efficient for the prediction of Tg of other glass systems with the multi-stage characteristic.展开更多
基金supported by the National Key R&D Program of China (2017YFB0701500 and 2017YFB0701600)the National Natural Science Foundation of China (51602187, U1630134, 11874254 and 51622207)the Shanghai Municipal Education Commission (14ZZ099 and QD2015028)
文摘Despite the usage of both experimental and topological methods, realizing a rapid and accurate measurement of the onset temperature(Tg ) of GexSe1-xglass transition remains an open challenge. In this paper, a predictive model for the Tg in GexSe1-xglass system is presented by a machine learning method named feature selection based two-stage support vector regression(FSTS-SVR). Firstly, Pearson correlation coefficient(PCC) is used to select features highly correlated with Tg from the candidate features of GexSe1-x glass system. Secondly, in order to simulate the two-stage characteristic of Tg which is caused by structural variation with a turning point at x = 0.33 via the structural analysis, SVR is utilized to build predictive models for two stages separately and then the two achieved models are synthesized using a minimum error based model for Tg prediction. Compared with the topological and other methods based on SVR, the FSTS-SVR gives the highest predictive accuracy with the root mean square error(RMSE) and mean absolute percentage error(MAPE) of 10.64 K and 2.38%, respectively. This method is also expected to be more efficient for the prediction of Tg of other glass systems with the multi-stage characteristic.
