Sign in

Graduate Research Assistant

Newark, DE
January 13, 2020

Contact this candidate


Prosper Kosi Anyidoho

Newark DE 302-***-**** Github Linkedln Kaggle


Research Experience: 2-year graduate research experience solving analytical problems through data collection, data processing, and developing machine learning algorithms. Programming & Tools: 2-year experience in Python, R, SQL, SAS, C++, Tableau, GIS, Hadoop, Spark, TensorFlow, Keras. Proficient with Data Analysis in Excel. Certifications: Coursera Deep Learning Certificates (4 courses) EDUCATION

University of Delaware

Master of Science, Civil Infrastructure Systems Aug. 2018-May 2020(Expected)

• Selected Coursework: Java Programming, Machine Learning, Advanced Data Analysis, Convex Optimization, Data Structures in C++, Data Analysis with SAS, Geographic Information Systems, Risk Analysis, Big Data Technologies, Applied Multivariate Statistics, Database Systems.

Kwame Nkrumah University of Science and Technology, Kumasi, Ghana Bachelor of Science, Civil Engineering Aug. 2013-May 2017

• Selected Coursework: Differential Equations, Algebra, Multivariate Calculus, Numerical Analysis, Statistics for Engineers, Data Analysis in Excel, Microeconomics, Macroeconomics.


University of Delaware August 2018 – Present

Graduate Research Assistant

• Data Acquisition: Deployed a web survey created in Qualtrics to collect data on household evacuation decision making during hurricanes. Also collected data on hurricane characteristics in GIS.

• Data Processing and Feature Engineering: Conducted data cleaning, visualization, and generated new features using feature engineering on Hurricane Data.

Kwame Nkrumah University of Science and Technology Aug. 2017- Aug.2018 Undergraduate Teaching Assistant

• Tutorials for undergraduate students in Algebra, Differential Equations and Data Analysis.


• Built machine learning classifiers with Hurricane Evacuation data to predict evacuation demand with accuracy as evaluation metric and improved predictive power on existing models by 5%. code

• Dynamic Discrete Choice Model in C++ to predict departure time decisions during Hurricane Evacuation (In progress). code


• A data-driven and Monte-Carlo risk analysis framework for train derailments in California: Safety and Policy Implications (Conference Paper Accepted).

Contact this candidate