Morgan Oneka

NYC Based Data Scientist

About

Hey! I'm Morgan, and I'm a Data Scientist based in the NYC metro area. Most recently, I was the customer data expert at Bloomingdale's; before this, I spent four years conducting bioinformatics data analysis at the University of Michigan.

I'm passionate about helping others make sense of their data, whether that involves flexing my modeling skills or simply running a few SQL queries.

In my spare time, I like playing music (I play piano, guitar/bass, and am currently learning violin!), creative writing, crochet, and quilting.

Skills

Python

  • Basics: numpy, pandas
  • Machine Learning: scikit-learn, catboost, lightgbm, xgboost
  • Web Scraping: BeautifulSoup, Selenium
  • Visualization: matplotlib, plot.ly, altair, bokeh

R

  • Visualization: ggplot, ggpubr
  • Spatial data: spatstat
  • Graphs/networks: igraph, tidygraph/ggraph

Web Development

  • HTML/CSS, Bootstrap
  • JavaScript, D3.js, visualization libraries like APEX Charts and PlotLy.js

Other Languages

SQL, LaTeX, SAS, MATLAB, C/C#/C++, Java

Data Science

  • Hypothesis testing, A/B testing
  • Model productionalization, monitoring, dashboards
  • Feature engineering, model selection
  • Git/GitLab

Domain Knowledge

  • Ads, retail
  • Cancer and molecular biology
  • Ecology

Portfolio

Here are a few projects I think highlight my skills. To learn more about other projects I've worked on, see my full portfolio page.

Customer Lifetime Value

Created in collaboration with coworkers during my time at Bloomingdale's.

  • Created a tree-based regression model using lightgbm to predict customer spend
  • Collaborated with data engineering team to productionalize using Airflow

DiNeR (Differential Network Visualization in R)

Created as part of research workflow during my time at the University of Michigan

  • R-based visualization library for small-scale network comparison
  • Designed to work with spatial transcriptomics datasets
  • Used to generate figures for several papers and grant applications
  • Code is available here

Haircare Product Usage

Personal project

  • Simple recommendation system for haircare products based on hair type
  • Dataset was obtained by scraping Reddit comments and Amazon reviews
  • Code is available here