Opened 3 years ago

Closed 3 years ago

#2137 closed defect (done)

The calculated stdev of the noise for feature selection problems is wrong

Reported by: mkommend Owned by: gkronber
Priority: medium Milestone: HeuristicLab 3.3.10
Component: Problems.Instances Version: 3.3.9
Keywords: Cc:

Description

Currently the sigma is calculated as sigma(target) * sqrt(noiseRatio). However, the correct formula is sigma(target) * sqrt(noiseRatio / 1-noiseRatio) to achieve a max R² of 1 - noiseRatio.

Change History (5)

comment:1 Changed 3 years ago by gkronber

  • Status changed from new to accepted

comment:2 Changed 3 years ago by gkronber

r10440: fixed a bug in the calculation of noise std.dev. in the feasture selection problem instances for symbolic regression

comment:3 Changed 3 years ago by gkronber

  • Owner changed from gkronber to mkommend
  • Status changed from accepted to reviewing

comment:4 Changed 3 years ago by mkommend

  • Owner changed from mkommend to gkronber
  • Status changed from reviewing to readytorelease

Changes in r10440 looks fine and the ticket has no dependencies => it can be released.

comment:5 Changed 3 years ago by gkronber

  • Resolution set to done
  • Status changed from readytorelease to closed

r10478: released bug fix for (#2137 The calculated stdev of the noise for feature selection problems is wrong)

Note: See TracTickets for help on using tickets.