Machine Learning Notes

Machine Learning review notes Classic Machine Learning Models KNN Classification and regression Non-parametric Algorithm: Given a vector $v$, calculate the distance between it and every vector in the training data Sort the distances descendingly, keep the smallest $k$ samples For classification, the prediction of $v$ is the most common class labels in the $k$ neighbors. For regression, the prediction is the mean value of the neighbors. Applications: anomaly detection, search, recommende...

Submit