Why does ensemble model work better than individual model?

A single model is known to make errors. If we make classifiers that make different errors, the misclassified examples will not be the same. Thus, the models can complement each other, and the error can be reduced. So, instead of training one model for the given problem, we can train...