Skip to content
View yejinPARK48's full-sized avatar

Block or report yejinPARK48

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
yejinPARK48/README.md

🌟 Nice to meet you!

I recently earned my Master's in Applied Statistics from the University of Michigan, and I'm currently a Research Assistant at the Institute for Social Research (ISR) and a researcher at the Statistics Online Computational Resource (SOCR).

My work spans poverty estimation in Michigan using large-scale survey data and fMRI analysis of music perception and brain activity using kime theory. My interests are multivariate analysis, graphical models, and statistical modeling for complex, high-dimensional data.

Here, I have summarized the projects I have worked on and am currently working on.


🛠️ Tech Stack

Frequently Used:

Also Familiar With:


📂 Data-Analysis-Projects

Here’s a collection of several projects I worked on during my undergraduate studies and now as a master's student. The projects are in R, Python, and SAS files. They are also linked to my personal website.

  • Python programming

    • Building Detection from Aerial Imagery for Survey Research: Built an end-to-end pipeline that detects and counts buildings from NAIP aerial imagery using a U-Net segmentation model (ResNet34 encoder, PyTorch), validated against Microsoft Building Footprints (tile-level MAE 1.42, Pearson r 0.95). Applied the model across 9 Michigan PUMAs to produce block-group-level building estimates for small area estimation (SAE) linking to ACS survey data, at the University of Michigan ISR (NCSES-funded).

    • Fine-Tuning BERT for Fake News Detection: Developed a transformer-based BERT(Bidirectional Encoder Representations from Transformers) model that efficiently classified real and fake news articles, leveraging a dataset of 72,134 news titles and applying key techniques such as preprocessing, text cleaning, tokenization, padding, and hyperparameter tuning.

  • R programming

    • Statistical ERGM analysis for consulting company network data: Examined the factors influencing advice-giving relationships within a consulting company's workplace network using the Exponential Random Graph Model (ERGM) to provide insights for improving workplace network efficiency and communication. Code

    • Analyzing HR Data using Machine Learning: Analyzed HR data from 14,999 employees to identify the key factors influencing turnover using various machine learning algorithms such as Logistic Regression, Decision Trees, Support Vector Machines, Random Forests, and Artificial Neural Networks.

    • Predictive Machine Learning Models for Diabetes Onset: Analyzed BRFSS survey data from 253,680 Americans to predict diabetes using machine learning models including ANN, Logistic Regression, and Decision Trees, integrating socio-economic and physiological factors.

  • SAS programming

    • Predicting the Index Trend of 'KOSPI' through Time Series Analysis: Predicted trends in the Korea Composite Stock Price Index (KOSPI) using search volume data from 2016 to 2021, and construted an ARIMA model.

Contact

Email GitHub LinkedIn

Pinned Loading

  1. michigan_building_detection michigan_building_detection Public

    Building detection from NAIP aerial imagery using a U-Net model, with PUMA- and block-group-level estimation for survey research.

    Jupyter Notebook

  2. ERGM_Company_data ERGM_Company_data Public

    Park YJ, Um JM, Hong SB, Han YJ, and Kim JH* (2022). Statistical ERGM Analysis for Consulting Company Network Data, The Korean Journal of Applied Statistics, 35(4) 527-541.

    R

  3. VerifyNews_using_BERT VerifyNews_using_BERT Public

    This project aims to detect fake news using BERT-based models leveraging a dataset of 72,134 news titles.

    Jupyter Notebook 1

  4. Analyzing-HR-Data-using-Machine-Learning Analyzing-HR-Data-using-Machine-Learning Public

    Y. Park, B Choi, Y. Woo, S. Yoo, and S. Kim, “Analyzing HR Data using Machine Learning”, Compilation of final projects for multivariate and big data analysis practical

    Jupyter Notebook

  5. Predictive-Models-for-Diabetes Predictive-Models-for-Diabetes Public

    R 1