Data Python

Tae Yoon Lim

917-***-**** *** W ** th St, New York, NY, 10024 ******@********.***

Education

Columbia University Aug. 2019 – Dec. 2020

Master of Science, Data Science

• GPA: 3.77/4.0

• Relevant Coursework: Machine Learning, Applied Deep Learning, Stat Inference & modeling, Computer Systems University of California, Los Angeles Dec. 2018

Bachelor of Science, Mathematics/Economics

• GPA: 3.54/4.0

Skills

• NLP, Computer Vision, Deep Learning (CNN, RNN, RCNN), Machine Learning

• Python (Pytorch, tensorflow, sklearn, pyplot), R, SQL, Spark, MQL5, C++, VBA Projects

Peace Speech Project https://github.com/tylim9307/peace-speech-project Sep. 2020 – Dec. 2020

• Analyzed news articles using NLP techniques, and study the relationship between the language used in articles and peacefulness of the country. Extracted a set of lexicons with Deep Learning and word frequency analysis, improved peace metrics Stock Movement Prediction https://github.com/tylim9307/Stock_prediction_through_DeepLearning Sep. 2020 – Dec. 2020

• Used LSTM, transformers, Attention, VLSTM to predict the movement of stock price, achieved 67% accuracy on movement Experience

Korea Investment & Securities Seoul, South Korea Jun. 2020 – Aug.2020 Quantitative Research Intern

My team publish automated analysts’ report by analyzing the news articles’ sentiment and level of importance, my task is to build a model to analyze the news in English that covers S&P500, and summarize important articles

• Achieved 4% increase in the accuracy of sentiment analysis of financial news article by fine-tuning BERT model with a linear classifier. Github: https://github.com/tylim9307/Text-Classification_sentiment-Analysis_financial-news-data

• Built Text Classification model by fine tuning pre-trained model (BERT, FinBERT) with a CNN layer on top

• Created a pipeline to summarize news articles using the fine-tuned distilBERT summarizer which has reduced the time spent of analysts on labeling the sentiment and importance of articles. 93% accuracy on labels

• Created a pipeline to web-scrape news articles on S&P500 with Selenium and preprocess text for the further research IRIMinds Seoul, South Korea Apr.2019 – Jul. 2019 Quantitative Research Intern

• Built python model to find pairs on KOSPI market based on the cointegration and perform the pairs trading; formulation period of 24 months and trading period of 12 months, backtested for 5 time periods

• Built Gold & Dead Cross model on MQL5 platform with selected pairs, adjusted parameters to achieve stable return

• Managed risks on algorithm by adjusting lot size, reducing slippage, and changing level of closing position Quarterback Investment Seoul, South Korea Jun. 2017 – Aug. 2017 Robo-Advisory Algorithm Intern

• Collected and pre-process data through Bloomberg and Thomson Reuters for the Robo-Advisory Algorithm

• Analyzed time series data of 32 nations’ benchmark stock indices and net difference between labor force and capital accumulation with Pandas and Numpy module in Python to find the correlation between indices

• Utilized VBA to automatically update data in Excel files and transformed the data into a usable form for Python UBS Seoul, South Korea Jul.2016 – Sep 2016

Equity Research Intern

• Assisted analysts by organizing and accumulating data useful for their reports

Contact this candidate