- Trying to pick out some new music, using data instead of a jukebox. Source: Patrick Fore, unsplash.com*
I want to create a recommendation system
- Introduction
- README Outline
- Repo Contents
- Libraries and Prerequisites
- Feature and Definitions
- Models
- Conclusions
- Future Work
- Built With, Contributors, Authors, Acknowledgments
Artist by popularity. I don't really care about pop music, but this graphic was a good way to learn more about the data.
This repo contains the following:
- README.md - this is where you are now!
- Recommender_System.ipynb - the Jupyter Notebook containing the finalized code for this project.
- LICENSE.md - the required license information.
- Data - test.csv and train.csv files
- CONTRIBUTING.md
- Images - contains the fun images to this repo.
These are the libraries that I used in this project.
- import pandas as pd
- import numpy as np
- import matplotlib.pyplot as plt
- %matplotlib inline
- import seaborn as sns
Just thinking about choosing music. Source: Natalie Cardona, unsplash.com
Not sure I will even have models.
In the end I really couldn't do it. I couldn't figure it out. I tried to use some examples that other people came up with for recommender systems but shy of copying the code I couldn't get them to work. Sometimes, even when I copied the code it still didn't give me the right answer. I know I want a Content-based system now.
I remember when this was a thing! Source: Brentt Jordan, unsplash.com
I think I need to do a few more basic recommender models - like the classic Movie dataset, or random numbers from the Surprise toolkit. I also need to read more about Content-based vs Collaborative models.
Jupyter Notebook Python 3.0 scikit.learn
Please read CONTRIBUTING.md for details
Thomas Whipple
Please read LICENSE.md for details
Thank you to Kaggle for the fun competition and interesting data: