Evaluation and Optimization of Leave-one-out Cross-validation for the Lasso
This work addresses a computational bottleneck for researchers and practitioners using lasso regression, offering incremental improvements in efficiency and scalability.
The paper tackles the problem of efficiently computing leave-one-out cross-validation for the lasso by developing an algorithm that produces a piecewise quadratic function, enabling exact hyperparameter optimization and demonstrating practicality on real-world data sets.
I develop an algorithm to produce the piecewise quadratic that computes leave-one-out cross-validation for the lasso as a function of its hyperparameter. The algorithm can be used to find exact hyperparameters that optimize leave-one-out cross-validation either globally or locally, and its practicality is demonstrated on real-world data sets. I also show how the algorithm can be modified to compute approximate leave-one-out cross-validation, making it suitable for larger data sets.