Trading Systems Labs says machine learning and genetic programming does work for quant and market forecasting
Someone form the vendor of
Comment from a successful GENETIC
Our LAIMGP is fast, accurate, efficient, writes code, explores wide ranges of solutions for directional, pairs, portfolios, HFT and other approaches and is rated #1 by Futures Truth on sequestered data in several categories beating a huge array of strategies written by worldwide Quants. Strategy developers and traders worldwide report live trading success with this technology in numerous markets and exchanges. So to say that GP does not work is actually quite inaccurate and uninformed.
Rapid idea testing is exactly one of the areas that Machine Learning can expand upon by using fact sets in novel ways that manual code writing simply cannot explore given the limits of human designs.
The ML design paradigm is now converging with the data and execution packages creating a higher level of engineering capability all controlled by the knowledgeable Quant. This will enhance and supplement coding and reward those who will evolve with it.
Also, This thread is all over the map with Lisp and SVM. We use GP not to predict but to create expectation. Perhaps that is the difference.
Any opinions out there?FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!