In a globalized world we live in today consequences of catastrophic events easily transgress national borders. Whether it’s a natural disaster, a war or an economic crisis it’s likely to spread out and affect all of us. We propose a framework to model global risks that is not bound to any specific model and is a hybrid of human and machine intelligence. The core of this approach is in using Bayesian Nets of causalities constructed by an analyst equipped with text mining and a map of economic, political and business interconnections.

We update probabilities of events happening conditional on priors based on the latest information releases. Any model can be plugged into the framework for propagation of events down to the target variables. Particular emphasis is on contagion effects that are consequences of interconnections network properties.

One selected candidate will receive a stipend via the DSI Scholars program. Amount is subject to available funding.

Faculty Advisor

  • Professor Eugene Neduv
  • Department/School: IEOR/SEAS
  • Location: Morningside Campus
  • Research Interests: complex system in Economical and Financial domains where graph analytics can be applied to understand the structure of such systems. The goal is to find concentration o systemic risk and use unsupervised learning to classify parts of the system.

Project timeline

  • Earliest starting date: 03/02/2019
  • End date: 08/30/2019
  • Number of hours per week of research expected during Spring 2019: ~20
  • Number of hours per week of research expected during Summer 2019: ~40

Candidate requirements

  • Skill sets: Familiar with Bayesian learning and machine learning, Natural Language Processing. Good programming skills are a must. Python, SQL and other databases languages. Need to have coursework in mathematics, economics, finance. Political science or social science courses are a plus.
  • Student eligibility (as of Spring 2019): freshman, sophomore, junior, senior, master’s
  • International students on F1 or J1 visa: eligible
  • Other comments: Need to have keen interest in machine learning and economics, political science.