Visit project site at http://vehicle-aging.herokuapp.com
Part 2: When Do Old Cars Become Cool Again?
Python (BeautifulSoup, Pandas, Sci-kit Learn, Plotly Dash)
The idea for this project came from a phenomenon observed with certain cars. New cars are valuable, but depreciate every year. However, after a certain number of years, some cars start regaining value again, presumably because people consider them classic.
I wanted to observed the value curves for various cars to determine when they become cool again, and to visualize how this varies by car.
For a car to be included in this analysis, I decided that it had to have been produced in at least one ‘old’ decade (60s, 70s, 80s, 90s) and at least one ‘new’ decade (00s, 10s, 20).
To find out which cars qualified, I scraped the Wikipedia lists of Cars Made in Decade for each decade since 1960. Then, I determined which cars had satisfied the above requirement.
Then, I web-scraped listings for each car from a used car listing website.
Next, I used Sci-kit learn‘s GridSearchCV and Support Vector Regression features to find the best regression parameters of asking price on model year for each car.
Finally, I built an interactive dashboard in Dash for comparing the value curves of different cars.
I think C4 Corvettes are awesome, but — like 90s models of many other cars examined — they appear to be the least valued version of its brand.
