When addressing a linear regression problem with an exceptionally large dataset in Spark ML, which method is most effectively utilized? Select the single best answer.