Averaging LASSO coefficients for repeated random partitioning of data


New Member
Is it reasonable to average LASSO coefficients from repeated reshuffling of training/test sets?

Suppose I randomly divide my data into testing & training sets, then within the training set use 10-fold cross-validation to choose an optimal λ, then refit on full training data and record the model coefficients. Now, suppose that I repeat this process k number of times. Each iteration will choose slightly different coefficients. One might think to average these together. However, each iteration may not choose the same set non-zero coefficients, therefore the average of all coefficients may contain many more non-zero coefficients than any single solution.