News for hackers. Stuff that matters.
Re:Fitting to Noise or Nothing at All: Machine Learning in M... by lettergram on Monday August 07, 2017 @03:49PM (#14950663)
For Piglet, I'm actually going back and conducting research to find valuable signals and releasing everything in as clear a way as possible.

Importantly, I'm providing access to signals first with explanation(s) of what each signal is / how it's useful today.

I also have an investment model currently accomplishing on average 50%+ YoY (for the back testing on 8 years, live for 4 years).

From there, I'm building further machine learning models on top of the signals. However, (as mentioned prior from the sentiment analysis paper) I'm doing much more robust research. For example, comparing all stocks as opposed to DJIA.

There are no comments on this comment yet.