Implemented Kmeans - Githubissues

ruokun-niu commented 1 year ago

Kmeans is a relatively easy-to-implement algorithm and I believe that it is a great starting point. Only numpy was used and no new external imports are needed. Please follow these instructions for setting up environment.

This PR tracks issue #18. Functions Implemented:

KMeans(): The core functionalities. Take in a dataset and an integer value of k. Returns a centroid of a Kmeans grouping and the datapoints that are associated with the centroid
biKmeans(): A variation of Kmeans that resolves the local minimum problem
randCent(): helper function for creating a centroid
distEuclid(): Calculates the euclidean distance between two vectors

Tests will be added soon (maybe in a separate PR)

JackyXu-Cool commented 1 year ago

LGTM. Good job Tommy!

ruokun-niu commented 1 year ago

Restoring the branch for assignment purposes

mmmmarty commented 1 year ago

Reviewed and great job

ruokun-niu commented 1 year ago

@hloneal Yep I agree. I'm gonna ask our client about using scipy for ML models during our next meeting.

For this algorithm, the client did provide some existing code so it's all right

JackyXu-Cool / Team-2130-Machine-Learning-Roulette

Implemented Kmeans #19