Hello,
While learning about ensemble learning I came to know that: to create a best ensemble model the local search from many different starting points may provide a better approximation to the true unknown function than any of the individual classifiers.
So for example if I am using random Forests in R,after I have run my model once how do I build an ensemble of models from three different starting points.Will setting the seed to three different values and then creating a train and test data work or is there some other method to follow?
Can somebody please help me on this!!