Machine Learning: Regression with GraphLab Create
biasvariance tradeoff  gradient descent  
Ridge regression  crossvalidation  measure of fit + measure of model complexity  
Lasso regression  coordinate descent  feature selection  measure of fit + (different) measure of model complexity  
Nearest Neighbor Regression & Kernel Regression  
concave  hax Max value  
convex  has Min value  
Models

• Linear regression • Regularization: Ridge (L2), Lasso (L1) • Nearest neighbor and kernel regression 
Algorithms  • Gradient descent • Coordinate descent 
Concepts

• Loss functions, biasvariance tradeoﬀ,
crossvalidation, sparsity, overfitting, model selection, feature selection 