Updates 16-04-2018

[rahlk]

More results

Remarks:

• Tuned learners perform better than the state-of-the-art (CoGEE)

  Others

• MOEA/D and NSGA-II still running
• Working on sharelatex project
• Graphics in progress ...

[arennax]

For DE tuning, we use 2 variants: DE10 and DE30. DE30 follows the rule that #np=#decision5, in our case its `6 3 = 30` populations DE10 uses the fixed number 10 as population size follows Wei Fu's work, which suggests that a not too large fixed number is enough for some case.

For bi-objective methods, the two objectives are: 1. minimize MRE; 2. Minimize the Confidence Interval associated to MRE. Both NSGA-II and original GA have population size = 100, as suggested in literatures.

Default #Gen = 250 with stop rules we discussed.

A larger Feature Model Tree is under constructing..