Hi All, How is stacking of different algorithms for prediction/ classification problems done ? If I intend to stack different classifiers in R , how it can be done ? If it can be explained with a sample R code. Does it leads to increase accuracy in prediction. Thanks in anticipation
Stacking, Blending and and Stacked Generalization are all the same thing with different names. Initially, when we have multiple classifiers to train the model then we use different techniques to combine the output of each classifiers like through voting, weighted voting, averaging the results, etc. This is the traditional way of ensemble learning.
In stacking, output of the classifiers (lower level classifier) will be used as training data for another classifier (Higher level classifier) to approximate the same target function. Basically, you let the Level 1 classifier to figure out the combining mechanism.