Portfolio

Education

• MA in Quantitative Methods in the Social Sciences (Concentration: Data Science), Columbia University, Graduate School of Arts and Sciences (GSAS)
• MPA in International Affairs (Specializations: United Nations Studies, International Conflict Resolution), Columbia University, School of International & Public Affairs (SIPA)
• BA in Political Science (Concentration: International Affairs), Northwestern University

Work Experience

• Data Scientist Intern @ Volkswagen Group of America (VGoA) - Belmont, CA - 2022-2023
• SIPA Capstone Project Consultant @ International Peace Institute (IPI) - New York, NY - 2020
• Protection of Civilians (PoC) Team Intern, Division of Policy, Evaluation, and Training @ United Nations Department of Peace Operations (UN DPO) - New York, NY - 2019
• Refugee Status Determination (RSD) Intern @ United Nations High Commissioner for Refugees (UNHCR) - Cairo, Egypt - 2016

Portfolio Projects

• GIS & Spatial Analysis:
  • Title: "Examining Geospatial Covariate Relationships with Civilian Killings in South Sudan’s Civil War"
    • Focus: Geospatial Statistics Technology Used: QGIS, R, GeoDa
    • Contents: Geospatial interpolation, hot spot analysis, Moran's I calculation, Moran's I residual analysis, Lagrange Multiplier (LM) and Robust LM lag and error diagnostics, and Spatial Durbin models.
• Data Visualization:
  • Title: "UNAMID: Did the UN’s Withdrawal from Darfur Lead to More Violence against Civilians?"
  • Version showing all code
  • Version not showing code
    • Focus: GIS Mapping
    • Technology Used: R Markdown, ggmap, tmap, sp, sf, rgdal, stadia/stamen maps, plotly, dplyr, ggplot2, ggthemes, ggpubr, stringr, scales, and kableExtra
    • Contents: Interactive and static charts, GIS maps, heat map tables, density maps, and union buffers and using statistics within them
• Natural Language Processing (NLP):
  • Title: "Using News Articles on Events in Nigeria in 2019 to Predict Political Violence"
    • Focus: NLP for Time-Series Forecasting
    • Technology Used: Python, TF-IDF word transformation, NLTK library, Scikit-Learn machine learning models, Scikit-Learn's TimeSeriesSplit, Augmented Dickey-Fuller Test, first-differencing, lags
    • Contents: Converting news articles by publishing date into time-series machine elarning forecasting models. Performance comparison between Ridge, Lasso, Random Forest, and XGBoost regression models
  • Title: "LDA Topic Modeling & VADER Sentiment Analysis for Political News Articles on Events Related to Nigeria in 2019"
  • Primary document (Python)
  • Visualization for the project (R)
    • Focus: Topic Modeling & Sentiment Analysis
    • Technology Used: Python, R Markdown, Excel, NLTK for stopwords, PorterStemmer, and PunktSentenceTokenizer, gensim library for CoherenceModel, LdaModel, and corpora, Jaccard similarity, vaderSentiment library, itertools, ggplot2
    • Contents: Text data cleaning, Latent Dirichlet Allocation (LDA) topic modeling of Nigerian news article text, VADER (Valence Aware Dictionary for Sentiment Reasoning) sentiment analysis scores for articles containing specific political words, compared across quarters of the year.