What to do once we have found an optimal model??
well the obvious answer is to tweak it and make it from a good model to an amazing model.
How to achieve it well stack 2 good model on another.Its like group A creates a product which is good and group B adds finishing touches to it and makes it look even better, thus adding value to the project.
I have made use of the all categorical UCI adult data set.
My 2 simple and best predicting models are naive bayes and Decision Tree. so I have stacked both of them on each other .
I failed to improve the accuracy or prediction powers in my example , but you could use it as a reference of stacking.
In my future articles i will show methods on how we can make the best use of stacking.
For the code and method please visit my GitHub link below