Thanks for the feedback, Yogita.

The concept of Delta AIC is basically a way to interpret the value -(AIC1-AIC2), where AIC1 < AIC2, in the following equation:

You can use the guidelines prescribed by Burnham and Anderson (2002) to interpret the coefficient -(AIC1-AIC2) i.e. (AIC2-AIC1). Or you can use the above equation of relative likelihood directly, to determine how much more probable is model 2 as compared model 1, in achieving the best possible goodness-of-fit on the data set in question. I like to use the relative likelihood equation, rather than the rules of Delta AIC, but it’s a matter of personal choice.

‘best