Comparing the logistic and hinge losses
In this exercise you'll create a plot of the logistic and hinge losses using their mathematical expressions, which are provided to you.
The loss function diagram from the video is shown on the right.
This exercise is part of the course
Linear Classifiers in Python
Exercise instructions
- Evaluate the
log_loss()andhinge_loss()functions at the grid points so that they are plotted.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Mathematical functions for logistic and hinge losses
def log_loss(raw_model_output):
return np.log(1+np.exp(-raw_model_output))
def hinge_loss(raw_model_output):
return np.maximum(0,1-raw_model_output)
# Create a grid of values and plot
grid = np.linspace(-2,2,1000)
plt.plot(grid, ____, label='logistic')
plt.plot(grid, ____, label='hinge')
plt.legend()
plt.show()