Yanmin (Mike) Ma, mathematics/economics major, and Manchen (Mercy) Fang, electrical and computer engineering/computer science major, spent ten weeks studying historical archives and building a model to predict the price of pigs, relative to a number of interesting factors.
Project Results
Berkshire pig transaction price is influenced by factors including the age of the pig, the number of offspring and the prevailing pork price. And we can conclude that within 1910 – 1920, breeders and buyers did price Berkshire pig rationally to a large extent.
Download the executive summary (PDF).
Disciplines Involved
- Economics
- History
Project Team
Undergraduates: Yanmin (Mike) Ma, mathematics/economics major, and Manchen (Mercy) Fang, electrical and computer engineering/computer science major
Client: Gabriel Rosenberg, Assistant Professor, Women’s Studies Program
Graduate student mentor: Chris Glynn, Department of Statistical Science
“Working with Mercy Fang and Mike Ma has been a pleasure. Throughout the Data+ experience, Mike and Mercy demonstrated creativity, diligence, and technical ability in their research on Berkshire pigs. Our primary objective was to build a data set that would cross-reference genealogical, market, and geographic data on individual Berkshire pigs. To collect this data, Mike and Mercy worked with original scans of archived primary source publications. Together, they developed efficient algorithms and error-correcting mechanisms to extract data from PDF documents. Mike and Mercy taught me a lot about working with text-data, the perils of optical character recognition, and the factors that drove hog markets in the early 1900s.
The pride that they took in their research is evident in their Data+ seminar talks and their final poster. Most impressively, Mike and Mercy were extremely independent and self-reliant. They formulated their own questions, developed strategies for answering them, and implemented solutions on their own. They are very promising young researchers.” — Chris Glynn, graduate student mentor
