class: center, middle, inverse, title-slide # Lab 08 ## Logistic Regression ### 04.02.20 --- ### Lab grading - **Completion** - **Complete**: Reasonable attempt at every question, work submitted as PDF, document, neatly organized with appropriate headings (50 pts) - **Partially Complete**: Reasonable attempt at most questions, work submitted as PDF, document, neatly organized with appropriate headings (25 pts) - **Incomplete**: Most questions unanswered or partially answered (0 pts) - **Formatting** - Work not submitted as PDF (-5 pts) - Document not neatly organized, unclear or missing exercise headers (-5 pts) --- ### Tip for remote collaboration: GitHub issues - Each GitHub repo has a tab called "Issues" that can be used to keep track of what needs to be done on a given project. - GitHub issues are especially useful when working with a team remotely, - One way to use issues on lab assignments and the project: - Create a new issue in the repo that includes a list of tasks for the assignment/project, - Assign each task to a team member by tagging their GitHub username (e.g. @Username) - Learn more about creating a new issue [here](https://help.github.com/en/github/managing-your-work-on-github/creating-an-issue). --- ### Lab 08 Introduction <font class="vocab">Goal: </font> Fit a logistic regression model to predict the probability a Spotify user will like a song <font class="vocab">Data: </font> Song characteristics from Spotify.com and like/dislike from a Spotify user - Link to the data dictionary are on the [Spotify documentation page](https://developer.spotify.com/documentation/web-api/reference/tracks/get-audio-features/) - Need a reminder on the code? See [logistic.html](https://www2.stat.duke.edu/courses/Spring20/sta210.001/appex/logistic.html)