The following notebook contains KNN implemetation from scratch on three different datasets.
It contains 2 different distance measures for KNN implementation which includes Manhattan and Euclidean distance.
It also contains mfcc feature extraction for speech data and bag of words representation for textual data.
The following blog explains the notebook in detail:
https://medium.com/@talhatahir_77835/knn-from-scratch-a-deep-dive-into-k-nearest-neighbours-9d061b3107fb