KNN Classifier

K-Nearest Neighbors (KNN) classifies a data point based on the majority class of its nearest neighbors. It finds the K closest training samples and assigns the most common class among them.

This simple yet effective algorithm requires no training phase. It memorizes the data and makes decisions at prediction time.

• KNN is sensitive to the scale of features. Always use the Feature Scaler node to normalize your data before applying KNN. Use the K-Finder analysis node to determine the optimal K value.
• Ideal for initial exploration of classification problems before trying more complex models.

Use Cases

• Room similarity detection: Find rooms most similar in shape and area to a reference room.
• Zone classification: Assign zones (residential, commercial, industrial) based on spatial properties.