【實用教育app】MSDMobile|最夯免費app

This app is part of the Northeastern CS6240 final project which deals with MSD on Amazon PDS by using map reduce.

It supports song year prediction and song recommendation.

First the features are extracted from the original 280GB dataset from the entire MSD.

Year prediction is a supervise learning problem, a NaiveBayes model is trained from these features.

Song recommendation is a easy clustering problem, we use k-means to get all the possible clusters.

The app's backend is a django server hosted on EC2, it grabs the data from S3 bucket in which the map reduce jobs output their data.

免費玩MSDMobile APP玩免費

免費玩MSDMobile App

MSDMobile APP LOGO

MSDMobile APP QRCode