Fast algorithms for fitting L_1-penalized multivariate linear models to structured high-throughput data

12/15/2017
by   Jane W. Liang, et al.
0

We present fast methods for fitting sparse multivariate linear models to high-throughput data. We induce model sparsity using an L_1 penalty and consider the case when the response matrix and the covariate matrices are large. Standard methods for estimation of these penalized regression models fail if the problem is converted to the corresponding univariate regression problem, motivating our fast estimation algorithms that utilize the structure of the model. We evaluate our method's performance on simulated data and two Arabidopsis genetic datasets with multivariate responses. Our algorithms have been implemented in the Julia programming language and are available at https://bitbucket.org/jwliang/mlm_packages.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset