If I want to carry out innovative research based on Wasserstein Regression, what other perspectives can I carry out statistical innovation? Wasserstein Regressions can I carry out statistical innovation? Specifically,(1) Combining with Bayesian framework, the prior distribution is introduced and parameter estimation is performed based on Bayesian rule to obtain more reliable estimation results.(2)Variable selection technique is introduced to automatically select the predictive distribution that has explanatory power to the response distribution to obtain sparse interpretation.
Can the above questions be regarded as a highly innovative research direction?