The following are some of my formal, public-facing projects:
Professional Reports & Software
A Tale of Two Cities: Alameda and Alameda Point
Peter Amerkhanian (2023)
Consultant Report for The City of Alameda Department of Finance.
Report Executive Summary
Summary: I utilize various demographic and financial data to analyze redevelopment alternatives in the Alameda Point region.
Tools: geopandas, sklearn.
A Reexamination of Proposition 13 Using Parcel Level Data
Peter Amerkhanian, Max Zhang, and James Hawkins (2023)
Report for The UC Berkeley Institute for Young Americans.
Report Executive Summary Data Appendix
Summary: We utilize 12 million property records in California to estimate the tax discount effects of Proposition 13 across property types.
Tools: tidyverse, marginaleffects.
Supporting Whole Families: SparkPoint® Community Schools
Peter Amerkhanian and Ena Yasuhara Li (2021)
Report for United Way Bay Area.
Report
Summary: I utilize client outcome data and interviews with clients and program staff to evaluate United Way Bay Area’s Community Schools program.
Tools: pandas.
Grade Migration Manager
Peter Amerkhanian (2019)
Software developed for Ismael Perez Pazmiño High School via The U.S. Peace Corps.
Code Slides Demo Video
Summary: I developed a full stack software application to migrate student grade data from a school’s excel-based system into the education ministry’s java application.
Tools: pyautogui, xlwings.
Academic Miscelanea
The following are some of my course papers from graduate school:
Simulating School Desegregation in San Francisco
Peter Amerkhanian (2022), Public Policy 275 Final Paper.
Paper Repo
Summary: I use a synthetic dataset of San Francisco public high school students and spatial optimization methods to simulate the effects of various busing strategies for racial desegregation outcomes.
Tools: geopandas, networkx. Methods: dijkstra’s algorithm, entropy/dissimilarity statistics.
Measuring Differences in California Politician Agendas in Press Releases
Peter Amerkhanian (2021), Information 254 Final Paper.
Paper
Summary: I use a novel dataset of press releases issued by governors and mayors in California to 1.) Develop a regression model to identify press release authorship, 2.) Cluster press releases by topic, and 3.) Estimate the political similarities between mayors and governors.
Tools: gensim, sklearn. Methods: logistic regression, latent dirichlet allocation.