Improving Linear Regression by Reducing Residual Errors

View
Here, we discuss the basic underpinnings of "boosting" and "gradient boosting". Focus on improving the estimation of a target function by computing "residuals" or errors– differences between the predicted and actual. If you follow the concept, it will become easier to use this idea to improve accuracy in linear regression, for example, by using the gradient boosting options available in many libraries.


Source: Andreas Müller, https://www.youtube.com/watch?v=OC3qmxGh2gc
Creative Commons License This work is licensed under a Creative Commons Attribution 3.0 License.