Sample Work

This repository is a small sample of some work I've done both in my free time as as part of my MS in Data Science. For additional information, you can always connect with me on LinkedIn.

Current Project

I'm currently working on a marketing campaign analysis. We have five months of data, with a marketing campaign being held in the third month. The goal of the analysis is to determine the efficacy of the campaign and identify next steps/areas of improvement.

I'm completing the analysis in two parts. The first part is just a general analysis: identifying a goal, reviewing the data, and getting a general overview of the results for the full client set as well as

Marketing Campaign Results - Part 1

  • Tools: Python, Jupyter Notebook, Pandas, NumPy, Plotly
  • Methods: Data exploration, merging multiple data sources, feature engineering, data visualization.
  • Summary: Using data provided from StrataScratch, I completed an initial analysis on a marketing campaigns efficacy including the impact of the campaign on different customer segments.

Past Projects

These are some of my favorite projects that I've completed. In some cases I'm still working on formatting and styling, and will likely be updating them over time. But for right now my priority is wrapping up my current project.

Collaborative Projects

Traffic Analysis

  • Tools: Python, Jupyter Notebook, Pandas, NumPy, Matplotlib, Seaborn, SKLearn
  • Methods: Data exploration, merging multiple data sources, data visualization, subsetting, feature engineering, automated multivariate feature selection, correlation matrixes, test/training splits, grid search, confusion matrixes, hyperparameter tuning, decision tree analysis, logistic regression, k-nearest neighbors, stacked ensemble, model scoring.
  • Summary: Using publicly available traffic data from the city of Portland, OR, my partners and I developed an analysis of key predictors for speeding behavior in the Portland Metro area. My role focused on the data exploration, data source merging, and final model generation.

EV Charging Optimization

  • Tools: R, RMarkdown, lessR, ggplot, Kable, ROI, OMPR, dplyr
  • Methods: Data exploration, outlier identification & removal, data visualization, model formulation & optimization
  • Summary: Using sanitized data provided by Portland State University's Transportation & Parking department, my team and I created an optimization model for installing new EV chargers on campus to meet current and future demands, including an alternative solution that may have greater benefit. My role involved formulating and implementing both optimization models as well as variable development.

Individual Projects

Fraudulent Account Detection

  • Tools: Python, Jupyter Notebook, Pandas, NumPy, Matplotlib, Seaborn, SKLearn
  • Methods: Data exploration, merging multiple data sources, feature engineering, VIF, multivariate feature selection, correlation matrixes, test/training split, hyperparameter tuning, data scaling, k-nearest-neighbors, model scoring.
  • Summary: Using data provided from StrataScratch, I developed a model to predict fraudulently created banking accounts.

AB Testing Game Changes

  • Tools: Python, Jupyter Notebook, Pandas, NumPy, Matplotlib, Seaborn, SciPy, Plotly
  • Methods: Data exploration, outlier management, statistical testing/t-test, data visualization
  • Summary: Using publicly available data I analyzed the impact on player retention of changing a key componant to an online game.

Database Updates

  • Tools: MySQL
  • Methods: Stored procedures, temporal database updates
  • Summary: A small example of managing updates to a customer database to maintain records while aging out inactive accounts.

Housing Price Prediction

  • Tools: R, RMarkdown, lessR, dplyr, tidyr, stringr, glmnet, Kable
  • Methods: Data exploration, data visualization, correlation matrixes, logistic regression, hypothesis testing, confidence intervals, model fit, colllinearity evaluation, prediction intervals.
  • Summary: Using data from Kaggle, I conducted a modely statistical analysis of housing features & prices to generate a model that predicts a house's price based on the selected features.

HBR Accounting Report Analyses

  • Tools: Word, Excel
  • Methods: Activity based costing, single cost drivers, billing models, pooled cost drivers, salary calculations
  • Summary: These are a few examples of some accounting reports generated from Harvard Business Review documents. I'm mostly including this for more of an example of my writing style than anything else.