Hello All!
Linear Regression is the key to understanding regression in machine learning. It is the algorithm where everything is born, and everything ends. It may not be as powerful as the Boosting algorithms are, but regression is the way forward since it has great interpretability.
While performing regression, two important metrics have to be noted and kept handy, for understanding the importance of variables, and how they affect the target:
How important is the predictor, for predicting the Target?
(R-squared will denote this)Was this predictor's importance by chance?
(P-value obtained on running the algorithm, will denote this)
Thanks for reading! Let's discuss it!
Top comments (0)