navjot12 / KNN_Analysis

This jupyter python script analyses performance of K-Nearest Neighbours algorithm with different distance functions. See final code cell for summary of results.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

#KNN Analytics

This python script compares speed and accuracy of K-Nearest Neighbour algorithm with different distance functions.

The algorithm has been applied on MNIST data- a collection of 42000 handwritten digit (0-9) images.

The final cell of the jupyter notebook provides a summary of the result.

Inference: Weighted K-Nearest Neighbour algorithm with Hellinger Distance provides the best accuracy. Time difference among methods was found to be negligible.

About

This jupyter python script analyses performance of K-Nearest Neighbours algorithm with different distance functions. See final code cell for summary of results.


Languages

Language:Jupyter Notebook 100.0%