The Machine Learning “Advent Calendar” Day 1: k-NN Regressor in Excel

The k-Nearest Neighbors (k-NN) regressor algorithm is explained through a practical implementation in Excel. It demonstrates the core principle of finding the 'k' most similar data points to a new observation and averaging their target values to make a prediction. The process is first illustrated with a single feature, covering distance calculation, neighbor identification, and the effect of the hyperparameter 'k' on overfitting and underfitting. The explanation then extends to using two features, introducing the Euclidean distance and highlighting the critical importance of feature scaling to prevent one variable from dominating the distance calculation.