Python Idle
MSE? MAE? Cross-entropy? Every neural network needs a loss function — it's the signal that tells the model how wrong it is. But different loss functions punish different kinds of errors. MSE hammers outliers. MAE treats all errors equally. Cross-entropy cares about confidence. Today you'll implement all three from scratch and see how each one shapes what the model learns.
~15 minsandbox lab
Loading Python runtime...
Goals: 5 tests
mse_value should be a float
mae_value should be a float
bce_value should be a float
MSE and MAE should give different values (they are different loss functions)
should create a matplotlib plot
Python loading...