Visualizing model complexity
Now you have calculated the accuracy of the KNN model on the training and test sets using various values of n_neighbors
, you can create a model complexity curve to visualize how performance changes as the model becomes less complex!
The variables neighbors
, train_accuracies
, and test_accuracies
, which you generated in the previous exercise, have all been preloaded for you. You will plot the results to aid in finding the optimal number of neighbors for your model.
This exercise is part of the course
Supervised Learning with scikit-learn
Exercise instructions
- Add a title
"KNN: Varying Number of Neighbors"
. - Plot the
.values()
method oftrain_accuracies
on the y-axis againstneighbors
on the x-axis, with a label of"Training Accuracy"
. - Plot the
.values()
method oftest_accuracies
on the y-axis againstneighbors
on the x-axis, with a label of"Testing Accuracy"
. - Display the plot.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Add a title
plt.title("____")
# Plot training accuracies
plt.plot(____, ____, label="____")
# Plot test accuracies
plt.plot(____, ____, label="____")
plt.legend()
plt.xlabel("Number of Neighbors")
plt.ylabel("Accuracy")
# Display the plot
____