, Seongnam, South Korea (Jan 2018 – Mar 2018)
Developed and wrote paper on a click model that predicts user click behavior, given search query and corresponding search engine results page. Utilized Generative Adversarial Network to precisely replicate real user behavior, especially incorporating sequential GAN, conditional GAN, and reinforcement learning, using Python (TensorFlow).
• Undergraduate Research Intern
, Data Mining Lab
, Seoul National University (Nov 2016 – Aug 2017)
Advisor: Professor U Kang
Researched on Deep Learning by writing papers on applying neural network on coupled matrix factorization (first author), and feature learning in signed directed networks (third author). Both were submitted, while SIDE: Feature Learning in Signed Directed Networks was accepted to WWW’18. Developed a time series data based stock price prediction project using deep learning. Modeled a neural network model that makes long term predictions (stock price after one to four quarters) on whether an individual stock price will rise, fall, or stay constant, which achieved up to 70.64% precision. Used Python (TensorFlow, NumPy, Pandas, SKLearn) for all implementations.
• Undergraduate Intern
, Software Platform Lab
, Seoul National University (Jul 2016 – Aug 2016)
Advisor: Professor Byung-Gon Chun
Developed a deep learning framework of Dolphin, a machine learning platform built on top of Apache REEF.
Used Java to build layers needed in neural networks and test benches.
, Care Rights
, Seoul, South Korea (Aug 2015 – Aug 2016)
Supervisor: Professor Sooyeon Han
Care Rights is a United Nations DPI/NGO(ECOSOC) nonprofit organization working for older person’s human rights. Contributed to legalizing “Act on Decisions on Life-Sustaining Treatment for Patients in Hospice and Palliative Care or at the End of Life” (Act No. 14013). Wrote and translated official documents in English, Korean and contributed to thesis writing.