Model selection for collector efficiency of seaweed drier by using LASSO and multiple regression analysis using 8sc

Abstract

Seaweed is considered as an important product used in different food and non-food items all over the world, so it's drying pattern is also important for gaining the best quality. The current paper explains the factors effecting on collector efficiency in v-Groove Hybrid Solar Drier under the climatic conditions of Malaysia. The current study examined the main factors with their interaction terms effecting on collector efficiency. Five variables were taken in this study with one dependent and four independent variables. Multiple Regression was used up to third interaction level by considering 32 all possible models with four independent variables for dependent variable (collector efficiency). On each model of multiple regression, Multicollinearity test and coefficient test is performed. LASSO is used as a sparse regression analysis to see the significant contributed factor like X1 (time), X2 (inlet temperature), X3 (collector average temperature) and X4 (solar radiation). In the model, Comparison is made for both type of regression analysis. As a result, LASSO is providing more efficient model as compare to multiple linear regression analysis.