Hyperparameter Optimization#

In this project we revisit Detecting Forgery with k-NN and add hyperparameter optimization.

Load Data Set#

Task: Load the banknotes data set (cf. Detecting Forgery with k-NN). Drop the entropy column.

# your solution

Grid Search with Cross Validation#

Task: Create a \(k\)-NN model with Scikit-Learn’s KNeighborsClassifier. Use hyperparameter optimization based on accuracy for choosing \(k\) and to find appropriate weights (uniform or inverse distance). Evaluate the model on a test set. Print optimal parameters and accuracy on the test set.

Solution:

# your solution

Decision Surface#

Task: Train a second model based on variance and skewness only. Plot the decision surface (the surface separating the classes) with Matplotlib’s contour or contourf.

Solution:

# your solution