Estimating conditional probabilities of historical migrations in the transatlantic slave trade using kriging and Markov decision process models

摘要: Intra-African conflicts during the collapse of kingdom Oyo from 1817 to 1836 resulted in enslavement an estimated 121,000 people who were then transported coastal ports via complex trade networks and loaded onto slave ships destined for Americas. Historians have a good record where these went across Atlantic, but little is known about individuals or enslaved \textit{within} Africa. In this work, we develop novel two-step statistical approach describe given documented violent conflict, transport peoples their location capture port departure, and---given individual's departure---that person's probability origin. We combine spatial prediction conflict density Kriging with Markov decision process characterising intra-African transportation. The results model can be visualised using interactive web application, plotting conditional probabilities historical migrations African diaspora. These help trace uncertain origins region Africa time period: methodology developed here provides probabilistic answer question.