© Hedged Capital LLC. All Rights Reserved.
Humans and machines have found it extremely difficult to predict asset price returns. Low signal to noise ratio and a continually changing trading environment are the two main reasons for this low level of success of both humans and machines.
We have successfully tackled this intractable prediction problem for over a decade and gained invaluable insights and experience into the theory and application of AI/ML technologies.
We welcome hard prediction problems in any industry. These projects lead to deeper insights into the applications of AI/ML algorithms to different industries, increasing our edge in trading the financial markets.
"Alas, it is always dangerous to prophesy, particularly about the future"
- Danish Proverb
We are honored that Google's TensorFlow Probability (TFP) Team has chosen our advisory services to design and develop educational materials about the application of TFP to computational problems in finance. TFP is Google's latest, open source, probabilistic machine learning library.
We are also excited about partnering with O'Reilly Media to develop educational content and cutting edge research in machine learning and computational finance.
Our first blog post and TFP model is about the 'Trinity Of Errors' inherent in all financial models and the urgent need to use probabilistic modeling and machine learning technologies to mitigate them. We are delighted that this blog post has received glowing reviews in the AI/data science communities.
Google's Josh Dillon, co-inventor of TFP, and Deepak recently spoke about Financial Machine Learning With TensorFlow Probability at the Trading Show West conference in San Francisco. You can download the presentation using this link.