Nearest Neighbour Classifier

• Find K nearest neighbours of a point, and classify the point on a majority vote of the neighbour classes
• point Similarity
  • use L2 Norm as distance
• larger the dataset, more robust the classifier
• brute force approach
  • calculate distance between the new point and every point in the classifier
  • computationally expensive