The Application of Machine Learning to Digitized Postcards as Artifacts

Abstract

When machine learning is mentioned in the news, it is rarely in the context of historical artifacts. Often, the public learns about what machine learning promises through news about optimizing workflows, improving trading algorithms, or even changing health outcomes for patients. However, this project demonstrates that machine learning techniques and tools can be used to develop new insights and collect new information on historical artifacts such as the Bowden Postcard Collection at Miami University.