Jennifer Nguyen
I deliver actionable data insights to support business decision-making through analytics.
MS in Business Analytics at George Washington University
Portfolio
Data Analytics
Best Model to Predict House Prices in Ames, Iowa
We want to predict house sale prices in Ames,IA using 80 housing features and to identify features that have the most influence on house sale price. The best model could be used by realtors and buyers to estimate price for a house with specific features that they want to sell/buy.
We applied 11 regression algorithms from linear to non-linear techniques, from parametric to non-parametric techniques. We conclude that the best model is the Gradient Boosting model with a test MSE of 0.0173, equal to test MAPE of 14.5%. Four most important predictors are Overall material and finish quality, Above grade (ground) living area square feet, Total square feet of basement area, and Original construction date.
Models: Best subset selection, LASSO, Ridge Regression, PCR, PLS, KNN, Regression Tree, Random Forest, Bagging, Boosting
We applied 11 regression algorithms from linear to non-linear techniques, from parametric to non-parametric techniques. We conclude that the best model is the Gradient Boosting model with a test MSE of 0.0173, equal to test MAPE of 14.5%. Four most important predictors are Overall material and finish quality, Above grade (ground) living area square feet, Total square feet of basement area, and Original construction date.
Models: Best subset selection, LASSO, Ridge Regression, PCR, PLS, KNN, Regression Tree, Random Forest, Bagging, Boosting
Forecast I-94 Westbound Traffic Volume for MN DoT ATR Station 301
The purpose of this project is to model the relationship between daily traffic volume (total number of vehicles) and time as well as between daily traffic volume and different weather conditions. Study results can be used to generate strategies, such as increasing tolls on specific weekdays, to alleviate traffic jams and help reduce car emissions.
We developed univariate deterministic & stochastic time series models and multivariate models (transfer function models). Predictors used in TF models are Average temperature, Amount of rain, and Average percentage of cloud cover. The best performing model achieves a test MAPE of 14.74.
Some interesting findings:
• Traffic volume is affected by the Average percentage of cloud cover;
• Friday is the busiest day of the week;
• Holidays are less busy than non-holidays
Models: Seasonal Dummies, Cyclical trend model, Error model, seasonal ARIMA, TF model
We developed univariate deterministic & stochastic time series models and multivariate models (transfer function models). Predictors used in TF models are Average temperature, Amount of rain, and Average percentage of cloud cover. The best performing model achieves a test MAPE of 14.74.
Some interesting findings:
• Traffic volume is affected by the Average percentage of cloud cover;
• Friday is the busiest day of the week;
• Holidays are less busy than non-holidays
Models: Seasonal Dummies, Cyclical trend model, Error model, seasonal ARIMA, TF model
Activities and Financial Situations of US Tax-exempt Organizations
This project will help us better understand the overall landscape of tax-exempt organizations in the Mid-Atlantic and Great Lakes region. The resulting database can be used as a directory to search for organizations of specific sizes, in specific locations, and with specific mission focuses, for example.
We also want to assess the financial situations of these organizations regarding profitability(income) and revenue. Financial health may disclose the quality of management and the preferences of American donors.
We cleaned and wrangled a data set with 640k+ rows from IRS into a star schema relational database using PostgreSQL and Linux shell commands.
Some interesting findings:
• The most popular activities for non-profits are Religion and Education, while the least popular is Advocacy (Weapon system, Racial integration);
• Geisinger Health Plan and Chancellor Masters & Scholars of the University of Oxford are the two not-for-profits that had the highest revenue in 2018;
• February and March are the two months with the highest number of organizations that first received exempt status
We also want to assess the financial situations of these organizations regarding profitability(income) and revenue. Financial health may disclose the quality of management and the preferences of American donors.
We cleaned and wrangled a data set with 640k+ rows from IRS into a star schema relational database using PostgreSQL and Linux shell commands.
Some interesting findings:
• The most popular activities for non-profits are Religion and Education, while the least popular is Advocacy (Weapon system, Racial integration);
• Geisinger Health Plan and Chancellor Masters & Scholars of the University of Oxford are the two not-for-profits that had the highest revenue in 2018;
• February and March are the two months with the highest number of organizations that first received exempt status
Consulting Projects
Pro-bono Consulting Project with Find Green
"Find Green is a start-up enterprise with a mission of empowering consumers to shift businesses toward sustainable practices by demonstrating market demand. Find Green has created the first consumer-choice platform (like Yelp, TripAdvisor) that allows users to find, compare and rate businesses based on their sustainability."
We created strategies to
• Increase engagement and retain current app users through data-driven incentive programs resulting from user survey analyses and secondary market research
• Identify partnership opportunities for client via research of market of sustainable businesses and green advocacy organizations
• Promote further product development
We created strategies to
• Increase engagement and retain current app users through data-driven incentive programs resulting from user survey analyses and secondary market research
• Identify partnership opportunities for client via research of market of sustainable businesses and green advocacy organizations
• Promote further product development
.jpg)
Market-entry Case for Gyrfalcon Ventures - Precision Analytics in Agriculture
Our client, Gyrfalcon Ventures, wants to exploit their recently-received ownership for international rights to GAAP, a platform of cutting-edge, market-leading precision agricultural analytics technologies that can be integrated into “quad-copter” drones. However, they are not sure which market to enter, while the value of GAAP is believed to diminish rapidly due to the client’s lack of incentive to invest heavily in R&D and actively innovating competitors.
We recommended that GV should enter Australia and a course of next steps to make the most out of this opportunity in a 3-year time window. To answer the question, we analyzed 15 metrics from World Bank and other reputable sources to rank 50 countries. Some of the notable metrics are Area of Average individual farm size, GPD per capita PPP, and strictness of drone regulation.
We recommended that GV should enter Australia and a course of next steps to make the most out of this opportunity in a 3-year time window. To answer the question, we analyzed 15 metrics from World Bank and other reputable sources to rank 50 countries. Some of the notable metrics are Area of Average individual farm size, GPD per capita PPP, and strictness of drone regulation.
