Efficient model selection of collector efficiency in solar dryer using hybrid of lasso and robust regression

Javaid Anam, . and Mohd. Tahir Ismail, . and Majahar Ali Majid Khan, . Efficient model selection of collector efficiency in solar dryer using hybrid of lasso and robust regression. pp. 193-210. ISSN 2231-8526

There are many variables involved in the real life problem so it is difficult to choose an efficient model out of all possible models relating to analytical factors. Interaction terms affecting the model also need to be addressed because of its vital role in the actual dataset. The current study focused on efficient model selection for collector efficiency of solar dryer. For this purpose collector efficiency of solar dryer was used as a dependent variable with time inlet temperature collector average temperature and solar radiation as independent variables. Hybrid of the least absolute shrinkage and selection operator (LASSO) and robust regression were proposed for the identification of efficient model selection. The comparison was made with the ordinary least square (OLS) after performing a multicollinearity and coefficient test and with a ridge regression analysis. The final selected model was obtained using eight selection criteria (8SC). To forecast the efficient model the mean absolute percentage error (MAPE) was used. As compared to other methods the proposed method provides a more efficient model with minimum MAPE.

Full text available from: Official URL: http://www.pertanika.upm.edu.my/resources/files/Pe...

Abstract

There are many variables involved in the real life problem so it is difficult to choose an efficient model out of all possible models relating to analytical factors. Interaction terms affecting the model also need to be addressed because of its vital role in the actual dataset. The current study focused on efficient model selection for collector efficiency of solar dryer. For this purpose collector efficiency of solar dryer was used as a dependent variable with time inlet temperature collector average temperature and solar radiation as independent variables. Hybrid of the least absolute shrinkage and selection operator (LASSO) and robust regression were proposed for the identification of efficient model selection. The comparison was made with the ordinary least square (OLS) after performing a multicollinearity and coefficient test and with a ridge regression analysis. The final selected model was obtained using eight selection criteria (8SC). To forecast the efficient model the mean absolute percentage error (MAPE) was used. As compared to other methods the proposed method provides a more efficient model with minimum MAPE.

[error in script]

Item Type:	Article
AGROVOC Term:	Solar drying
AGROVOC Term:	Hybrids
AGROVOC Term:	Models
AGROVOC Term:	Application methods
AGROVOC Term:	Data collection
AGROVOC Term:	Multicollinearity
AGROVOC Term:	Regression analysis
AGROVOC Term:	Solar radiation
AGROVOC Term:	Time
AGROVOC Term:	Prediction
Depositing User:	Mr. AFANDI ABDUL MALEK
Last Modified:	24 Apr 2025 00:55
URI:	http://webagris.upm.edu.my/id/eprint/9376

Actions (login required)

View Item