LinearRegression

linear-regression-workflow.R

A workflow for processing linear regression

1. Get correlations
2. Scatterplot of independent vs. dependent variable
3. Density plot of variable distributions; check for normal distribution to use linear regression
4. (Optional) Box plot to visualize outliers
5. Build linear model
6. Evaluate linear model: t-stat, p-value, goodness of fit with f-stat & standard error, adjusted r^2, AIC, BIC
7. Rebuild linear model with training & test sets; measure accuracy via correlation of actual vs predicted values
8. K-fold cross validation: cross-validate the linear model with k-folds; analyze