This is intended as a personal dumping ground for work produced while reading and following the book Introduction to Statistical Learning with applications in R (ISLR).
Top
A non-alphabetized glossary of the more technical terms and phrases in ISLR. Terms are in order of introduction in the book.
Mean squared error or MSE: How snugly the model fits the data. Training MSE applies to training data, and test MSE applies to test data.
Bias: The error that is introduced by using approximation.
Variance: How drastically f^ changes between different training data sets.
Bias and variance relate to model flexibility. As a model becomes more flexible, its variance increases and its bias decreases. That is, the model fits the training data more snugly.