p

g

,

ntly correlated with the time lag variable was examined as well.

Lasso regression model, the fitness was measured using the R-

easurement. Figure 7.23(a) shows the Lasso regression model for

sequences. In this Lasso model, top five 3-mers were ATT, TAT,

C and TAC. The R-square was 0.2345. Figure 7.23(b) shows the

gression model for the India sequences. The R-square for this

as 0.3209 and top five 3-mers were CAA, TTA, CTT, TTG, and

(a) (b)

he Lasso regression models constructed for modelling the relationship between

g and the 3-mer data. (a) The USA model. (b) The India model.

(a) (b)

he Lasso regression models constructed for modelling the relationship between

g and the 3-mer data. (a) The Russia model. (b) The Brazil model.

e 7.24(a) shows the Lasso regression model for the Russia

s. The R-square for this model was 0.5776 and top five 3-mers

A, TCA, CTT, CCT and CCA. Figure 7.24(b) shows the Lasso

n model for the Brazil sequences. The R-square for this model

51 and top five 3-mers were CCA, CCT, TCC, TAC and ACC. It