Background and Motivations
Online auctions have become a major part of the economy as well as an integral part of the consumer shopping experience.
The combination of the thrill of bidding/competition and the potential of getting a smart bargain drives consumers of all age towards ebay and other online auction sites.
In addition, it is in this online, secondary market place, that we witness psychological phenomenons such as shrouding and moral hazard and adverse selection.
In this project, we chose to explore the dynamics of Ebay's online auction system in part due to our combined interest in modeling, economics, and previous work experience.
We hope that by attempting to create models that predict final auction price based on time series and item characteristic data, we can showcase the tools we've learned in CS109.
Methodology
Although explained in more details in the subsequent pages, here is a quick summary of our analysis:
1. Scrapped over 200mb worth of auction data from ebay
2. Implemented 3 models to attempt to predict final price
3. Compared the results of 3 models and discussed limitations