is there any discussion or paper about how to define/design the labels for the ML training? Intuitively I can think of:

So in general: how to define your target variables (labels if classification) for stock predictions?

Thank you!

  • $\begingroup$ It almost seems a machine learning problem upon a machine learning problem: what is the best labelling or discrete labelling distribution to return the best test results from my original machine learning algorithm... $\endgroup$ – Attack68 Jan 19 at 17:20

A good beggining could be the paper of Gu, Kelly and Xiu (2018).

  • $\begingroup$ The paper greatly illustrates the features, but for labelling it simply uses the net return of the next time stamp. $\endgroup$ – mojovski Jan 22 at 22:44

I see there that you mentioned the text book Advances in Financial Machine Learning. At the back of every chapter are a list of good papers that provide some insight into the chapters body of knowledge. Chapter 3 is titled Labeling and it is of course a large part of the process.

I suggest reading all 39 papers in the bibliography. It was a real eye opener and they are well selected so you build a very deep understanding of the various techniques.

  • $\begingroup$ thank you! Will definitely take a look into this $\endgroup$ – mojovski Apr 3 at 20:55
  • $\begingroup$ Yeah, the papers show super basic stuff tipping in the dark. Simply applying svm, trees or similar old concepts on net returns between samples. My next steps will be to define the labels as a result of a entry/exit strategy such as from here: quora.com/What-are-day-trading-entry-and-exit-points $\endgroup$ – mojovski Apr 5 at 8:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.