• Built Social Media and News Based Trading Strategies with machine learning methods• Extract-Transform-load (ETL) user Data and Tweets with Twitter API tools and SQL database, scored different tweets from 0 to 1 by the rate, fit the sentiment score with different finance corpus weight.• Used Three NLP model (TruncatedSVD, TF-IDF, Word2vec) in python to train finance-related tweets, classified tweets by four factors (Accuracy, Precision, Recall, F1)• Backtested and find profitable… Show more • Built Social Media and News Based Trading Strategies with machine learning methods• Extract-Transform-load (ETL) user Data and Tweets with Twitter API tools and SQL database, scored different tweets from 0 to 1 by the rate, fit the sentiment score with different finance corpus weight.• Used Three NLP model (TruncatedSVD, TF-IDF, Word2vec) in python to train finance-related tweets, classified tweets by four factors (Accuracy, Precision, Recall, F1)• Backtested and find profitable strategies with 36.2% excess return and using SQL Server Analysis Services (SSAS) analyze how different factors have influenced the model profit Show less

• Built API tools environment and Select different users lists by age, Extracted young traders’ investment behavior data through Python scraper from China largest social media platform Weibo• Used Python Pandas and NumPy package Extract-Transform-load (ETL) data source as Time-series data • Built indices to model for calculating sentiment score using Python NLTK package, Used SQL database to store sentimental analysis score of social media • Defined sentimental index of trading and… Show more • Built API tools environment and Select different users lists by age, Extracted young traders’ investment behavior data through Python scraper from China largest social media platform Weibo• Used Python Pandas and NumPy package Extract-Transform-load (ETL) data source as Time-series data • Built indices to model for calculating sentiment score using Python NLTK package, Used SQL database to store sentimental analysis score of social media • Defined sentimental index of trading and conduct benchmark comparison, Optimized index-market trading system performance and increasing efficiency by analyzed and set different parameters• Visualized profits data with Tableau by creating a dashboard, analyzed economic indicators and reported the commercial feasibility of potential products Show less

• Created dashboards and analyzed market budgets and provide monthly PnL reports on clients• Retrieved, analyzed and attributed data by writing complex queries in SQL• Designed, maintained and delivered daily calculation of complex indices models via SQL, most of which are based on structured financial products and complex financial concepts• Created and ran statistical probability analysis using large data sets of financial returns, including correlation analysis and tail risk… Show more • Created dashboards and analyzed market budgets and provide monthly PnL reports on clients• Retrieved, analyzed and attributed data by writing complex queries in SQL• Designed, maintained and delivered daily calculation of complex indices models via SQL, most of which are based on structured financial products and complex financial concepts• Created and ran statistical probability analysis using large data sets of financial returns, including correlation analysis and tail risk analysis such as value-at-risk (VAR) under Basel III regulation Show less