Machine Learning - Ensemble Learning - Introduction to Ensemble Learning Tutorial

Ensemble methods is a machine learning technique that combines several base models in order to produce one optimal predictive model. Ensemble methods usually produces more accurate solutions than a single model would. 

Example 1] - KBC use audience poll where the participant will get more accurate result  from majority(i.e wisdom of crowd).

Example 2] – In online purchase, we are more confident in purchasing a product  having average rating greater than 4, than a product having single rating greater than 4. 

In ensemble, we can either use different algorithm with same data, OR same algorithm with different data OR different algorithm with different data to create variety in ensemble.

For Classification, we use max voting technique, and for Regression, we will use averaging technique.

Benefit-

• Improvement in performance
• Reduce Bias and variance
• Robustness

Each base model should have accuracy more than 50% atleast.