Comments (4)
Comments:
ws is just multiplying the matrixes, without subtracting the min
My mistake. I replaced W * Y
by W * (Y - minP)
in row 43 of scalarization_ws() to fix the problem.
wt is actually ws, without the Tchebycheff norm
The Tchebycheff norm is there! W * (Y - minP + eps)
does the Hadamard (i.e., element-wise) product, and applying max
by rows (with the apply(..., MARGIN = 1, FUN = max)
does the trick (since the Tchebycheff norm is equivalent to returning the largest value of the resulting row vectors).
awt is missing
Indeed
double check pbi calculations (seems to be missing a norm division in D2)
My mistake again. I replaced (Y - minP - D1 * W) ^ 2
by (Y - minP - D1 * W / NormW) ^ 2
to fix the problem.
all scalarization functions seem to be missing the min step which is included in the paper.
No. This is dealt with in the population update step. ;-)
from moeadr.
Hmmm, I forgot to protect the AWT from div/0. Will try to sneak that in today.
from moeadr.
HmHMmm.... The calculation of rho in AWT (page 4) requires the inverse of the weights, which means we need to deal with weights with zero values.
If we just add an epson to the zero values, the resulting Rho will be quite extreme. Example:
Weight Matrix:
0 0.25 0.25 0.25 0.25
0 0.25 0.25 0.25 0.25
0 0.5 0.125 0.25 0.125
0 0.125 0.125 0.25 0.5
Rho:
1 0 0 0 0
1 0 0 0 0
1 0 0 0 0
1 0 0 0 0
Is this acceptable?
from moeadr.
I believe it is. I'll consult Lucas about it, but from my understanding that should be the expected behavior - it certainly the is if we use an asymptotic argument.
from moeadr.
Related Issues (20)
- TODO: moead() output HOT 1
- WISHLIST HOT 1
- implement Violation Handling
- Evaluation stop criteria is not exact HOT 4
- Suggestions for more stop criterias HOT 1
- Modularize Neighborhood Assignment Functions
- Fix decomposition_uniform [PRIORITY] HOT 2
- Verify iPBI
- Verify calculation of the tchebycheff norm
- Case Study One
- Case Study Two
- Case Study Three
- Package "smoof" should not be necessary for build? HOT 1
- Found bug
- Preset_moead() -> MOEA/D-DE polynomial mutation parameters different from paper HOT 1
- Check Uniform Design HOT 2
- Different Initialization Methods
- Introduction of Discrete Variation Operators
- Manual initialisation + save/resume
- Add Riesz s-energy method for generating scalarisation vectors
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from moeadr.