Branch: refs/heads/master Home: https://github.com/scikit-learn/scikit-learn Commit: 370b642f7ae44d4157001d09d3a4cf1a2b990847 https://github.com/scikit-learn/scikit-learn/commit/370b642f7ae44d4157001d09... Author: Daniel Perry <daniel-perry@users.noreply.github.com> Date: 2019-07-02 (Tue, 02 Jul 2019) Changed paths: M doc/whats_new/v0.22.rst M sklearn/neural_network/multilayer_perceptron.py M sklearn/neural_network/tests/test_mlp.py Log Message: ----------- FIX Adding max_fun parameter to MLP with lbfgs solver (#9274) Since MLP estimators doesn't provide the derivative of the objective function to `fmin_l_bfgs_b` it needs to be manually evaluated, and so the number of function calls can be significantly larger than the number of iterations.