← Projects
Project · 2017

FeelTheMarket

My role
Hackathon project
Topics
machine learningmarket prediction
Links
FeelTheMarket

In 24 hours I explored the feasibility of a stock-market predictor driven by the sentiment of economic news and tweets: a simple TensorFlow regression model correlating past market data with news sentiment, used to test how different headlines would move a stock.

The project won the Bloomberg Prize at CopenHacks 2017, held at Microsoft HQ in Denmark.