ML Toolbox 3-Linear Regression using Scikit-Learn

Goal: 

• Ordinary Least Squares (OLS) Polynomial Regression Model
• Investigate bias-variance tradeoff (overfitting-underfitting) using the RMSE vs. degree curve 
• Stochastic Gradient Descent (SGD) Linear Regression Model

 

Reading Task:

• Polynomial Regression: Frequentist Approach Links to an external site. (bias-variance tradeoff, i.e., overfitting-underfitting, using the RMSE vs. degree curve). Note that this notebook uses Scikit-Learn's Ordinary Least Square (OLS) method Links to an external site., i.e., the closed-form solution approach.
• Gradient Descent for Linear Regression Links to an external site. (this notebook uses Scikit-Learn's Stochastic Gradient Descent (SGD) Links to an external site. algorithm)
• Polynomial Regression using SGD Links to an external site. (this notebook uses Scikit-Learn's Stochastic Gradient Descent (SGD) Links to an external site. algorithm)

Practice Task:

Implement the marked sections in the “ML-Toolbox-3-Linear-Regression Links to an external site.” Jupyter notebook.

The dataset is given in the file “winequality-red.csv Links to an external site.”.