Machine Learning & Data Analytics

Module 04 — Hit Songs, Case Study Introduction

Song Photo by Austin Neill on Unsplash

Introduction

You've been hired by a major record label, MMC Entertainment, to help them build a model that can predict whether or not a song will be a hit.

Stakeholders

These are the individuals your team will be helping during the case study:

Ezra, Lead Singer of the Wasps

I don't need some computer telling me how to make my art.

But, it could be interesting to incorporate some synthetic components into one of my songs...just as an experiment.

Tom Jones, Head of marketing and brand development

While we all celebrate the art behind your music Ezra, our research over the years tells us that there are several factors that go into predicting whether or not a song will be a hit.

Some of that comes down to timing, but we think there may be other aspects that we can quantify.

Johnny, Data Science Intern

Hey! You're here too! I just started working here last week.

The record company has a ton of information on old songs, including how popular they were. I suggest we start there.

Stakeholder Focus Areas

The stakeholders are particularly interested in the following areas:

  1. Building a model that can predict how popular a song is.
  2. Understanding how tastes have changed over time, and if there are cycles in those tastes.

You've been invited to a strategy meeting tomorrow. They're planning to discuss the marketing campaign and would like your input on a few key points.

Data

Spend some time with your team evaluating the data. Be sure to look at data types, ranges, and meanings of each feature from the data dictionary.

Datasets

You can use the following Google Colab notebook to assist you:

Open In Colab


  1. Lead Singer photo by Brian Lundquist on Unsplash 

  2. Head of Marketing photo by LinkedIn Sales Navigator on Unsplash 

  3. Data Science Intern photo by Fábio Lucas on Unsplash