The classical maximum likelihood estimation using Akaike's criteria is defined by $$\text{AIC}=-2\log^-\text{(maximum likelihood)} + 2 \text{(no. of independently adjusted parameters within the model)}$$
or mathematically, $$\text{AIC}[p]=N\ln(\sigma^2)+2p$$ where $N$ is the number of sampling and $\sigma^2$ the variance of the model when the order is set to $p$.
This method was introduced last 1974 almost 50 years and it has been cited 58,000 times.
Is there a latest method which is widely used and has been applied to many applications?