Analyzed feedbacks on Article Highlight model, provided visualized data summarization for tuning model

Built debugging tool for Article Highlight project to improve algorithm performance

Worked on training, deployment, knowledge transfer documentation of Article Highlight model

Worked on developing Doc2Vec, MLP and random forest classifier for Quick Reply project

Carried out validation on massive data in order to provide clients with good quality cash reports

Coordinated with Development Team, analyzed on bugs and improved the existing code for the application

Set up cash forecasting logic depending on historic payment rolls

Configured and compiled customization for new client’s application in QlikView for ETL process from ERP system or bank statements for visualization and KPI purposes

Extracted data from SQL Server, computed statistical summary, visualized data through Matplotlib

Modeled cumulative abnormal return (CAR) of classified stocks over benchmark within a 3 months’ time window through Numpy, Pandas, Datetime packages implemented in Python using OOP design

Developed program that auto-generated daily stock investment filter based on events (news, announcements etc.)

Detected positive and negative influence events out of 17 events, provided outline for reference and future updates

Conducted top-down analysis by comparing real estate markets in the West, North East and South East region of the US based on population, industrial structure, education and etc. The assembled data was then categorized into an Excel spreadsheet for investment evaluation