Hello,

Consider Bayesian analysis, penalized regression methods, principal components regression, partial least squares regression, and bootstrap methods to improve estimation reliability and prevent overfitting.

Hope this helps.

Hsin-Yuan Chen you are right your statement is supported by the finding of this article comparing ML and bayesian estimation .

Here you find the full article

Article Analyzing small data sets using Bayesian Estimation: The cas...

Jamal Tikouk

Thank you for sharing that article. It indeed reinforces the value of Bayesian estimation for small datasets.