Moisture Control Methods in Silk Reeling Process of Tobacco Based on the Random Forest Regression

Bo Ma

School of Statistics and Mathematics, Yunnan University of Finance and Economics, Kunming, China.

DOI: 10.4236/ojs.2015.55041                

Abstract

The moisture control of materials in silk reeling technology of tobacco is regarded as the key factor influencing the inner quality of cigarette. In this paper, according to the statistical data of the silk reeling production line of Yunyan (Ruanzhen brand) of Qujing cigarette factory from June 2013 to May 2014, it is feasible to apply the random forest regression model to study the problem of moisture control theoretically. In the perfuming stage of silk reeling, a random forest regression model is established to describe the change of moisture content of finished cut tobacco in the export link of perfuming stage, aroused by several factors including incoming water content and different environment. According to the model, good moisture control in the export link of perfuming stage (accordance with the technological standards) can be realized by adjusting the regulating reference value of incoming moisture under specific workshop environments. In the drying stage of silk reeling, the most effective method of moisture control is to adjust the cylinder wall temperature by means of analyzing the correlation coefficients among variables which influence the moisture content of cut tobacco in the export link of drying stage and then establishing another random forest regression model. And this method is consistent with the traditional production experiences. In conclusion, these methods referred above provide strong theoretical basis for stable moisture control in the export link of perfuming stage.

Keywords

Random Forests Regression, Moisture Control, Moisture Content in the Export Link of Drying Stage, Moisture Content in the Export Link of Perfuming Stage

Conflicts of Interest

The authors declare no conflicts of interest.

References

[1]	Breiman, L. (2001) Random Forests. Machine Learning, 45, 5-32. http://dx.doi.org/10.1023/A:1010933404324
[2]	Svetnik, V., Liaw, A., Tong, C., Culberson, J.C., Sheridan, R.P. and Feuston, B.P. (2003) Random Forest: A Classification and Regression Tool for Compound Classification and QSAR Modeling. Journal of Chemical Information and Computer Sciences, 43, 1947-1958. http://dx.doi.org/10.1021/ci034160g
[3]	Lin, Y. and Jeon, Y. (2006) Random Forests and Adaptive Nearest Neighbors. Journal of the American Statistical Association, 101, 578-590. http://dx.doi.org/10.1198/016214505000001230
[4]	Wu, X.Z. (2013) Statistical Methods for Complex Data—Based on R. China Renmin University Press Co., Ltd, Beijing, 33.
[5]	van der Laan, M.J. (2006) Statistical Inference for Variable Importance. International Journal of Biostatistics, 2, Article 2.
[6]	Peters, J., De Baets, B., Verhoest, N.E.C., Samson, R., Degroeve, S., De Becker, P. and Huybrechts, W. (2007) Random Forests as a Tool for Ecohydrological Distribution Modelling. Ecological Modelling, 207, 304-318. http://dx.doi.org/10.1016/j.ecolmodel.2007.05.011