* Apologize for Multiple Posting * * Please pass this Call for Partecipation to interested Colleagues * Int. Summer School on...
Mario Pavone
mpavone@...
Mar 1, 2008 9:32 pm
4459
Dear GP'ers, This is the 2nd and the last CFP of the special issue of the Journal Advances in Artificial Neural Systems. The topic is - Dynamics and Learning...
Dear GP'ers, I made a mistake in the previous announcement. <<2007>> should be <<2008>> Sorry for this but let me try once again This is the 2nd and the last...
* Apologize for Multiple Posting * * Please pass this Call for Partecipation to interested Colleagues * INTERNATIONAL SUMMER...
Giuseppe Nicosia
nicosia@...
Mar 5, 2008 4:05 pm
4462
We are seeking an outstanding postdoc in the field of machine learning, with experience / interest in at least some of the following topics: learning...
Juergen Schmidhuber
juergen@...
Mar 6, 2008 11:54 am
4463
Hi, For calculating diversity and for some other reasons I need a distance metric between two GP trees. The one I want is the edit distance between two trees....
Hi, I've been scouring the web for some papers related to this topic. The most promising seems to be this one.... http://www.mit.edu/~brossman/ted.pdf I've...
- FINAL CALL FOR PAPERS - FINAL CALL FOR PAPERS - WORKSHOP ON Evolutionary Computation and Multi-Agent Systems and Simulation Workshop (ECoMASS-2008) to...
I have recently been experimenting with a genetic algorithm to evolve multi-layer perceptrons to produce a particular output. The protoype output is just an...
parlos, ... "In this manner" could be taken in a couple of different ways, but are you familiar with the GP bibliography? It focuses on GP but also covers a...
Hi If your pictures are simply 32x32 pixels and grayscale, then you try to find the good gray level for 1024 pixels by using just one error measure. It's easy...
Hi, Have you already read about NEAT? It's an algorithm for evolving NN's. Also, I think it would be useful for you to read about MOGA (Mutation Only Genetic ...
... can be found at : http://www3.itu.edu.tr/~okerol/Mutation%20Operators.pdf This is a study of one of my students. You may well check other studies to my web...
I'll start with a question. you said you used a fitness test of euclidean distance in RGB values. This would instinctively be my first choice, What problems...
Hi Paros, We tried a multi-objective approach to the general problem of simultaneously optimizing both the neural architecture and minimizing errors. I refer...
... Would you mind describing the input space and maybe an example? ... Ditto. Do three labels mean anything related to the input? How many images are you...
Tom Osborn
osborn@...
Mar 9, 2008 3:14 am
4474
... euclidean distance in ... problems did you ... admit that i'm speculating here} ... The fitness metric used was each pixel contributes euclidean distance ...
Funny that you should mention flags. hxxp://home.comcast.net/~exspaceA/canadaH1M26.png Here we see the algorithm exploiting a "dirty trick" to get a ...
YOu know, I'd love to respond quickly to these quick little questions, but the moderators of this forum can take upwards of 15 hours to approve all my posts....
Our apologies if you receive multiple copies of this announcement. ... 11th International Workshop on Learning Classifier Systems to be held as part of the ...
Paros, You don't say what resolution image you're trying to fit. For hi-res images, it's very hard to do this with MLPs (or NEAT, for that matter), but more...
* Apologize for Multiple Posting * * Please pass this Call for Partecipation to interested Colleagues * Int. Summer School on...
Giuseppe Nicosia
nicosia@...
Mar 9, 2008 6:53 pm
4480
You might want to separate the R,G,B components out if you haven't already. If you can solve for any particular one, you can solve for all three, and there's...
Should you choose to go the distributed GA way as suggested below, the following might help you: "Promoting diversity using migration strategies in distributed...
I used deterministic crowding (DC) for years to combat this "premature convergence" problem and recommended it to colleagues and it was very easy to implement...
CALL FOR PAPERS International Workshop on Advanced Research Challenges in Financial Evolutionary Computing to be held as part of the 2008 GENETIC AND...
Christopher D. Clack
C.Clack@...
Mar 10, 2008 5:03 pm
4484
If I understand this right, you're trying to construct a 2-24-32-3 MLP that can more or less reconstruct a specific 80x80 image. If you're using a limited...
I intentionally avoid using anyone else's neural networks out of the box. My neurons have a transfer function of T(S) = S/(S+R) Where R is a constant giving...
Yes yes. I had the same intuitive idea back at the beginning the project. It turned out to be wrong! Here is a copy-pasted response I made to A W on this...
Thanks for the suggestions, but my final goal here was to test out this strange neural network that I'm using. It has "response constants" in each node and a...