Empirical risk minimization is simply the minimization of the loss on the training data given a set of hypotheses (e.g., logistic regression models).  You can implement it in Logistic in WEKA by setting the ridge parameter to zero. This will implement empirical risk minimization for logistic regression models for the log loss. (Note that the default ridge parameter is very small, so effectively Logistic implements empirical risk minimization by default.)

Minimizing the empirical risk doesn't make sense for unconstrained decision lists like those generated by PART because minimizing the error/loss on the training data will normally just overfit. Thus, PART uses regularization in the form of pruning. Note that PART implements a greedy heuristic algorithm for finding a small and accurate decision list. There isn't any mathematical guarantee that the PART model will be optimal in some sense.


On Thu, Dec 21, 2017 at 8:03 PM, Valerio jus <valeriojus@gmail.com> wrote:
Hi all, 

What approach can be used to find the "empirical risk" of a particular classifier in Weka.--how to find the empirical risk minimization in logistic regression, for instance? 

In sum, kindly I need the implementation of empirical risk Minimization on PART algorithm and logistic regression algorithm.

I do appreciate if any of Weka developers help me?


Wekalist mailing list
Send posts to: Wekalist@list.waikato.ac.nz
List info and subscription status: https://list.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html