Academic Projects
Welcome to my Academic Projects Portfolio! This is a space where I can showcase my academic projects in data science, consulting, and product management. Each of these projects allowed me to develop and implement various analytical and problem-solving skills. I hope you enjoy browsing through my work!
Data Science: Python
This project aimed to investigate the impact of Chronic Obstructive Pulmonary Disease (COPD) on adult mortality by employing Bayesian and Frequentist Generalized Linear Models (GLMs) to generate posterior distribution plots and linear regression results, and utilizing Random Forest Regression to quantify the influence of various elements on mortality.
The analysis revealed a high statistical confidence level of 90% - 95% for the reported relationships between COPD and adult mortality, providing robust evidence of their association.
Data Science: Python
This project was completed as part of work for my role as an Undergraduate Researcher at UC Berkeley Data Science.
The objective of this research project was to leverage data science and machine learning techniques to optimize fuel logistics networks and develop operational plans that effectively address associated risks.
The research was conducted in collaboration with the U.S. Air Force Operational Energy, utilizing various methodologies and tools such as Multiple Regression Models, Exploratory Data Analysis (EDA), Deep Learning, Neural Networks, Matplotlib, Seaborn Graphs, TensorFlow, Numpy, and Pandas.
Businss Case
LuMate is a company dedicated to enabling older adults to safely age in their own homes. With a significant portion of the elderly population living alone, LuMate recognizes the need to address the concerns of fall detection and prevention, as well as vital sign monitoring.
LuMate proposes a revolutionary solution that leverages radar technology, ensuring accuracy without compromising user privacy. Unlike existing wearables or camera-based sensors, LuMate's non-invasive software utilizes radar sensors, such as Google Nest Hub, to detect vital signs, falls, and prevent accidents.
By analyzing the collected data, LuMate generates valuable healthcare insights. With a projected North American market opportunity of $1,155 billion by 2029, LuMate aims to cater to adult children caregivers and independent older adults alike. Their product package includes a hardware radar kit and a software subscription, offering different plans to meet varying customer needs.
With an estimated gross margin of 83.7%, LuMate plans to allocate its funds towards product development, company operations, and marketing, with the capability to produce approximately 150,000 radar kits in the initial batch.
Businss Case
Background information: Recognizing the potential of the South Korean market, Oski Corporation (a prominent player in the US nutritional supplement market) conducted extensive market research and identified it as the most attractive market for their new product. The market research revealed a growing consumer awareness of the benefits of dietary supplements among young people in South Korea, coupled with a limited variety of nutritional products available in the country. Therefore, Oski has strategically decided to enter the South Korean market as their initial step, with plans to expand further into other countries in Asia in the future. The company intends to manufacture the nutritional pill in the US. It is exploring two options for market entry into Korea: either independently or through a partnership with a well-known Korean company.
​
After a holistic evaluation of financial statements and extensive research, we gave the following summary:
-
Expanding the Oski Corporation globally by first penetrating the South Korea market
-
Moving the needle in South Korea (if we choose partnership) through television marketing for brand populization
-
Creating international lines catering to cultural tastes and customer ideals
Businss Case
​
-
Participated in a project with DataGood to enhance the Food Bank of Contra Costa & Solana’s Interactive Map using Tableau.
-
The project aimed to utilize Census Data and identify Food Swamps.
-
The process involved understanding the data, cleaning and merging relevant columns from different datasets, handling missing values (latitude & longitude), and working with census tract data.
-
Data was collected from the 2020 Census (census.gov) and cleaned in Deepnote. It was then merged with shapefiles for census tracts and overlayed on Tableau.
-
Faced challenges in learning new technologies and software to fit technical needs, data collection, and combining and synthesizing Tableau worksheets.
-
The project provided valuable insights and recommendations, contributing to the understanding and mitigation of food insecurity.
​
The project involved working with datasets containing demographic information such as median household income, poverty status, language spoken, race, gender, age, percentage of renters and homeowners, and Internet access, among others. Additionally, we incorporated food distribution and food insecurity information, including data on food stamp recipients, food insecurity percentages, and food distribution patterns.
​
By conducting in-depth analyses on the dashboard, we were able to discover meaningful relationships and patterns within the data. This allowed us to gain insights into the factors contributing to food insecurity and the disparities faced by different communities. We utilized interactive filters to ensure that all the dashboard pages interacted seamlessly, allowing users to explore the data and understand the nuances of food insecurity in real time.
​
One of the key aspects of our project was ensuring that the map updated dynamically with new data, enhancing the accuracy and relevance of the application. By continuously updating the map with the latest food distribution and food insecurity data, we aimed to provide stakeholders with up-to-date information for decision-making and resource allocation.