Demonstrate designing and creating machine learning systems using best practices and patterns.
Recommendation Software
|
This program takes a brief user survey and uses it to recommend a song. To do this it uses a type of Collaborative Filtering algorithm which takes users that have similar likes and dislikes and makes suggestions based off that. I created all my own data for this project (100 songs, and 1000 ratings).
The plan for this project is to later be integrated into my SIP Project (the underground) to recommend bands and venues based off the users' preferences gathered from inside the app.
Check out the GITHUB here |
Beer Profile Prediction
For this prediction software, I started by collecting the color, abv%, and ibu numbers from 150 different types of beers (rows) to use as features. I then labeled them and ran them through pandas for processing and training.
Once the user enters the requested information about color, abv%, and IBU, the program will tell them what type of beer they could have based off the information I trained it on.
![]()
|
© JOE PARISIA, 2022 | ALL RIGHTS RESERVED