What is there inside a Trading Algorithm? Beállítások

[Michael Logan]

One of the topics I have written about here has been trading algorithms. These algorithms have had several problems considering outliers as “average” data and incorporating them as a new input in this learning process. I have always said that the good thing about machine learning statistical models is the “learning” part, but that there are times when they do not distinguish what they have to learn from what they have to let go. They do not distinguish between noise and signal. But this is solved by simply generating control flags to understand how the model performs in its behavior.