The Promise and Pitfalls AI and Machine Learning for Trading
Artificial Intelligence (AI) is a confusing term to the non-practitioner. It requires a discussion with each person to find out what they really mean. The details may be hidden in some goal to improve a rules base procedure being implemented by a programmer with no formal training in data science or machine learning. The details may be hidden in a research project where someone is trying to find a signal for a future event.
If AI is confusing then Machine Learning (ML) is baffling to the general public. When one reviews professional firms touting ML you find that they are often using simpler components including regression analysis, feature classification, optimization and probabilistic inference. These are really data science techniques that become the building blocks of ML. ML becomes a reality when these building blocks are used to create a predictive model that can produce a timely decision. ML is becoming a very real possibility in the world that now has Big Data, massive computing resources, and the data scientists to utilize both.
In the linked video, put together during the Security Traders Association of Chicago’s 92nd Annual Mid-Winter Meeting, the John Lothian News team discussed these confusing terms with several industry and academic veterans including our own Morgan Slade.
“The most promising area of Quantitative Trading is using it (AI & ML) … to identify alpha that you wouldn’t normally be able to find by eyeballing things and looking at regression statistics and visualizations,” said Slade.
Video courtesy of John Lothian News http://www.johnlothiannews.com/2018/03/myths-misconceptions-surrounding-artificial-intelligence/