One of my most fun grad school projects was a machine
learning system for predicting stock market indexes. It was a mix of sentiment
analysis, text mining, ensemble learning, and NLP for stock market prediction.
It was interesting to see how the models of major stock market indexes change
depending on what kind of news sources supplied data for the text mining
training sets. It was surprising (or not) that political data feeds had more
impact on the model than the info related to technical analysis/fundamentals
and traditional economics stuff. Today, these systems are called Robo Advisors
and appear to be increasingly popular.
https://www.financial-planning.com/news/robo-advisors-best-and-worst-returns-over-four-years
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.