• Extracted features from the Media Exposure and Opinion Formation (MEOF) survey.
• Processed large quantities (millions of rows) of user trace data from YouGov
• Utilized Python and R for proper data cleaning
• Segmented and documented code for modularized implementation
• Optimized Machine Learning models, neural networks, and utilized natural language processing for topic modeling

The model performance for ideology is significantly better.

GitHub Repository

Paper

• Machine Learning
• Data Processing
• Data Processing
• NLP
• Neural Network
• Governance