Hello! My name is Beomsu Kim, also known as YuKiSa / @shuuki4. I am a third year student, majoring in Computer Science & Engineering at Seoul National University.
I am highly interested in the field of machine learning and data science : including deep learning, NLP, optimization, etc.
Feel free to contact me.
I am an undergraduate student studying about Computer Science & Engineering at Seoul National University! My ultimate goal for studying computer science is to solve various kinds of challenging real-life problems. Thereby, I am highly interested in such fields - including Machine Learning, Data Science, Optimization, etc. Among these topics, my research interest is highly focused on Deep Neural Networks and applications of Deep Learning (NLP, CV). Furthermore, I'm occasionally participating in some Kaggle competitions. I love to learn new technologies and develop new ideas about challenging topics. I hope to develop skills as a data & machine learning scientist, and willing for chances to work in these fields.
서울대학교 컴퓨터공학부에 3학년으로 재학중입니다. 현재 머신 러닝, 데이터마이닝, 데이터 과학, 최적화 문제 등 쉽게 풀리지 않는 실생활의 다양한 문제들을 풀 수 있는 방법들에 대해 공부해나가고 있으며, 그 중에서도 딥러닝 및 이의 응용 분야 (NLP, CV)에 큰 관심을 가지고 있습니다. 데이터 과학 대회 사이트 Kaggle의 대회에도 간간히 참여하고 있습니다. 머신러닝 과학자, 데이터 과학자로서 다양한 경험을 쌓고 실력을 기르며, 이 분야에서 도전적인 연구를 하면서 일하게 될 수 있기를 기대하고 있습니다.
I have graduated Gyeonggi Science High School in Suwon, Korea.
I am now majoring in Computer Science & Engineering at Seoul National University, Seoul, Korea.
I have won the silver prize in the KOI(Korea Olympiad in Informatics) 2012 Algorithm Section.
I have wrote the graduation thesis for high school, about the push recovery algorithm of bipedal humanoids. I proposed a new modeling method for bipedal humanoids, and the policy-gradient based reinforcement learning algorithm for push recovery.
I have been receiving the Presidential Science Scholarship from Korea Scholarship Foundation.
I worked on a timetable generating program for Gyeonggi Science High School. Timetable generating problem in this school was post-enrollment based, including student sectioning problem. To solve the problem, we used FP-Tree and multi-stage metaheuristic search algorithms, and implemented a real-working program in Java.
I worked on the undergraduate research program in SNU IDS Lab. My research topic was Entity Linking for Tweets, especially focusing on entity resolution. In this work, I build & trained a LSTM-based network that approximates the joint probability for entity-tuple appearance using English Wikipedia Data, and combined it with simple LR model + SA technique to find the optimal answer for the given input.
I've participated in the 2016 Summer School Program at HEIG-VD . We studied various computer science subjects; including Machine Learning, Compiler Construction, Software Reverse Engineering, and Computer Networks.
I am currently working in Vision & Learning Lab as an undergraduate intern.