[Please reply DIRECTLY TO ME ([log in to unmask]) rather than spamming
Hi everybody. I truly apologize if you get multiple copies of this
I am working on a grant proposal to support further development,
improvement, and significant expansion of the ECJ toolkit. My goal is
to move ECJ towards a general-purpose, industrial-grade metaheuristics
toolkit. As part of this proposal I need to gather evidence of ECJ's
historical use over the last ten years, and particularly its current
use. Here are examples of evidence (there could be lots more):
- Use of ECJ in published or unpublished research
- Use of ECJ in education
- Use of ECJ in dissertations and theses of all kinds [VERY HELPFUL]
- Use of ECJ in commercial work [ALSO VERY HELPFUL]
- In-house use of ECJ of any kind
- Extensions to ECJ available online
- Tutorials or manuals about ECJ (other than my own)
For publications and theses etc. it'd be particularly helpful (but not
required -- I'll do the legwork) if you provided me with BibTeX entries.
Also I'd appreciate some pull quotes I could use describing any of the
- How you use the system
- Your opinion of the system
- How well we have provided support over the last ten years
Last, I'd like some opinions on directions we should take in extending
ECJ. Here are some I have in mind. Which matter most to you? What
am I missing that you'd really like to see?
- Better internal software testing (unit tests, etc.)
- Better reporting and parameter sweeping
- A usable GUI
- Frameworks for easier parallel computing
- More and better representations (Note: grammatical evolution, PUSH,
and maybe NEAT are on their way)
- More and better multiobjective optimization support (note: NSGA-II
is on the way).
- Ant colony optimization, GRASP, and other combinatorial methods
- Optimization with hard constraints
- Good single-state optimization support (hill-climbing, simulated
annealing, tabu search, ILS)
- Good memetic / hybrid algorithm support
- Diversity maintenance methods
- Heavily revised support for DE and PSO
- What else would you really like to see?
Which SPECIFIC algorithms and/or representations would you like to see
Thank you in advance for any help you can provide along these lines!
Anything is helpful.
George Mason University
[log in to unmask]