Fig. 3From: Feature selection for RNA cleavage efficiency at specific sites using the LASSO regression model in Arabidopsis thalianaOptimizing the tuning parameters in the LASSO regression model. Average mean square error (MSE) was calculated in tenfold cross validation (tenfold CV) by changing the parameter λ in LASSO regression. The number of non-zero features in the LASSO regression model was also calculated, and the parameter λ was determined according to the average MSE and the number of non-zero features in the LASSO regression model (a). The same data processing approach was used for the LASSO regression model, which used only sequence information (b). The Y-axis indicates the number of non-zero features in the LASSO regression model. The X-axis indicates the average MSE in tenfold CV. λ (0.01) was used in each regression modelBack to article page