For my final project at Metis, I decided to look at how Twitter responded to a controversial story. The story I picked came from May 30th, 2017 when TMZ leaked a photo in which Kathy Griffin was holding a mask that looked like Donald Trump’s bloody, severed head. My goal for this project was to see how people responded to something controversial and to get an understanding of how something controversial goes viral. For my presentation and for this blog post, I did this analysis in 3 acts.

For my third project, I looked at classifying LendingClub.com loans as either defaulted on or fully paid off. Lending Club is a crowdfunding loan website that allows borrowers to get loans at rates lower than banks, and investors to invest in those loans and get pretty good returns. I wanted to build a model from the perspective of an individual investor investing in loans.

For my second project, I looked at trying to predict the decline between week 1 and week 2 at the box office. When I was looking at movies on Box Office Mojo, I found most movies dropped pretty significantly from week 1 to week 2, but some movies increased, and I wanted to look at the factors behind the change. 

From now on we are on our own. Our projects from now on would be individual projects. This week we were introduced to our new project in which we need to scrape data and perform linear regression on that data. As of right now, I am choosing to predict the 2nd week box office performance for a movie. I got this idea, from noticing that most movies experience a significant decline in box office revenue from the first week to the second. However, a minority of movies have an increase. I want to look into the variables that are associated with this. I have scraped data from Boxofficemojo and pulled in api data from Omdbapi.

Our first project for Metis involved using subway data to try to find the best subway locations in NYC to spread awareness (through getting emails), and to get people to attend a gala supporting women in tech.

I have gotten through my first week of data science bootcamp at Metis and it has, so far, gone really well. It has been a lot of work, and we definitely have jumped right in, but overall I have not felt overwhelmed (yet). This week, we have covered basic pandas, python and git and focused on a group project involving NYC subway data.