k-Nearest Neighbors (k-NN) Classifier in Python using the numpy library. Creating a class function knn_classifier which takes as input: train_features (feature set of the training data), class_labels (associated class labels), test_features (features of the testing dataset), and num_neighbors (size of the neighborhood). Using numpy.linalg.norm and numpy.argsort to find the k closest neighbors, and using numpy.unique to determine the majority vote within a neighborhood.

Applying the Iris dataset from the sklearn library to evaluate the classifier with different k values. Measuring accuracy on both training and testing sets, using 67% of the data for training.

Additionally: an implemented k-fold cross-validation to optimize the choice of k. Dividing the training data into k equal segments, using k-1 folds for training and the remaining fold for validation. Repeating this process k times, each time using a different fold for validation. Calculating the average accuracy across all k folds to determine the optimal k value.