I'm a little confused on what exactly I'm seeing in picrel. Y axis is profit, X axis in the indicator. Is what's being shown that the higher the indicators value, the higher the likelihood of profit? Or am I going about this wrong way? This just seems like it's too simple. Although, I was expecting it to show the opposite.
linear regression with scikit-learn
- Thread starter barkbark
- Start date
You will probably want to calculate R^2 (https://en.wikipedia.org/wiki/Coefficient_of_determination) on that for a start... Spontaneously it looks like noise to me, and also the supposed largest magnitudes of "profit" occur in the middle of the parameter range.