Give sample dataset with at least five data point and N number of features on how those two below formulas can works: Empirical Risk Minimization

hat$(f)=argmi{n}_f{\displaystyle \underset{i=1}{\overset{N}{}}}l(f(x_i),y_i)$

Structural Risk Minimization

hat$(f)=argmi{n}_f{\displaystyle \underset{i=1}{\overset{N}{}}}l(f(x_i),y_i)+(f)$

$(f)$ is known as a penalty or regularization

term to control the model complexity of $f$

$>0$ is a trade$-$off hyper$-$parameter