Open arennax opened 6 years ago
-looking sane
questions:
Todo (in suggested order or priority, first to last):
Ideas:
Yes, I decide to use NP=50 to get the initial results sooner since 100 was too slow. Will add china and runtime
when will you add china and runtimes?
for our own GA, NP=50 is arguable but it could be said that at NP=100 we will beat DE much more often
in any case, when you do nsga-II and moea/D make sure you use their defaults. and if that is NP=100, then so be it.
but keep lives=5
I am running china now so can get result today, same as runtimes. roger for the defaults
For DE tuning, we use 2 variants: DE10 and DE30. DE30 follows the rule that #np=#decision*5; DE10 uses the fixed number 10 as population size follows Wei's work.
For bi-objective methods, the two objectives are: 1. minimize MRE; 2. Minimize the Confidence Interval associated to MRE
9 datasets, 3-fold cross validation, pop=50, gen=100, repeats=20
Experiments
For DE: (NP = 50, F = 1, CR = 0.5, life = 5)
Separate the data into train-part and test-part.
(Gen 0) Randomly generate 50 config (after constraints check), for each config[i] (i=1~50), calculate its mMRE (median MRE) on train-part.
(Gen 1~N) Use DE to generate 50 new config from precious Gen, and calculate their mMRE on train-part. For each config[i], if new config[i]'s mMRE is less than old config[i], use new config[i] to replace old config[i].
Stop rules:
Use config with least mMRE in Gen N, calculate its mMRE on test-part.
Since 20 repeats and 3-fold, we got 20*3 = 60 mMRE values for each dataset.
For GA: (NP = 50, CX = 0.6, MUT = 0.1, life = 5)
Separate the data into train-part and test-part.
(Gen 0) Randomly generate 50 config (after constraints check), for each config[i] (i=1~50), calculate its mMRE (median MRE) on train-part.
(Gen 1~N) Use GA to generate 50 new config from precious Gen, and calculate their mMRE on train-part.
Stop rules:
Use config with least mMRE in Gen N, calculate its mMRE on test-part.
Since 20 repeats and 3-fold, we got 20*3 = 60 mMRE values for each dataset.
Current Results (between ATLM, DE and GA):
A sorted graph between DE250 and GA250 in isbg10 dataset:
Runtime GA vs DE:
Number of Gen Comparison (between DE and GA):
Next Task
Add MOEA/D
Try NSGA-II with adjusted modification
Use DE/GA to tune CART
More literature review for potential paths
Update current OIL with uniform frameworks (DEAP/PyGMO..)
To Do
Re-construct OIL architecture (sklearn/utils/model/optimizer)
pip install package
Tutorial Materials (workshop to REU students)
Reverse negative results (Negative Results for Software Effort Estimation, 2016)