Regularization Programming assignment

Course Q&A Deep Learning Specialization Improving Deep Neural Networks: Hyperparameter tun
1

Hi everyone,

Does anyone know why the cost function after backward_propagation_with_regularization is increasing. I passed all the test so I’m assuming it supposed to be increasing but don’t know why.

2

Hi @Zahra_Bgh ,

If you are referring to running the model that uses:

  • compute_cost_with_regularization instead of compute_cost
  • backward_propagation_with_regularization instead of backward_propagation
    I could not see how the cost curve is increasing, in fact, it drops very quickly at round 1000 iterations and continue decreasing thereafter very slightly, almost staying low at round that level.