Which algorithm uses a specified K and a distance measure to classify a point by majority among its nearest neighbors?

Which algorithm uses a specified K and a distance measure to classify a point by majority among its nearest neighbors?

• A. Random Forests
• B. AdaBoost
• C. Support Vector Machines
• D. K-Nearest Neighbors

Answer: (D)

The idea is to classify a point by looking at its neighbors and taking a vote. You choose a number K and a distance metric. For a new instance, you compute the distance from it to every point in the training data, pick the K closest points, and assign the class that appears most often among those neighbors. The distance metric (like Euclidean or Manhattan) determines which points are considered nearest, so it shapes the decision boundary. The value of K trades off bias and variance: a small K makes the decision boundary flexible and sensitive to noise, while a larger K smooths it but may blur distinctions between classes. This approach is lazy and instance-based, since it doesn’t build a complex model during training but uses the stored data directly at prediction time.

The idea is to classify a point by looking at its neighbors and taking a vote. You choose a number K and a distance metric. For a new instance, you compute the distance from it to every point in the training data, pick the K closest points, and assign the class that appears most often among those neighbors. The distance metric (like Euclidean or Manhattan) determines which points are considered nearest, so it shapes the decision boundary. The value of K trades off bias and variance: a small K makes the decision boundary flexible and sensitive to noise, while a larger K smooths it but may blur distinctions between classes. This approach is lazy and instance-based, since it doesn’t build a complex model during training but uses the stored data directly at prediction time.