Developing novel methods of utilizing various data sources, ranging from macro economic indicators to news headlines, filtering noise, and creating algorithms to best project performance of a security or index over multiple time periods.
Using multiple data sources and the application of algorithmic techniques to determine the most profitable combinations and proportions of securities to hold while dynamically predicting necessary changes over time.
Sentiment Analysis on Financial Market News
Using Transformers for natural language processing with focus on sentiment calculation and analysis of financial market news. Analyzing how market sentiment affects market prices and how this can be used to shape a portfolio.