[annevolve] 20 ANNs fresh from the fat farm

[Dave Harris]

I ran the individual and multiple weight-zeroing process on 3-3s-4n-perf.sc. The individual zeroing is just for information purposes - no ANN adjustments were performed based on those numbers.

The multiple zeroing was pretty astonishing. One of the ANNs still functioned after zeroing 11 out of 28 weights.

I have attached the run report, the new sc files, and the source code for wbreduce (must be linked with fitness.c and neuron.c).

Dave

[Mitchell Timin]

Post by Dave Harris
I ran the individual and multiple weight-zeroing process on
3-3s-4n-perf.sc. The individual zeroing is just for information
purposes - no ANN adjustments were performed based on those numbers.
The multiple zeroing was pretty astonishing. One of the ANNs still
functioned after zeroing 11 out of 28 weights.
I have attached the run report, the new sc files, and the source code
for wbreduce (must be linked with fitness.c and neuron.c).

That's pretty cool, Dave. I'm pleasantly surprised that you were able
to make an ANN needing only 17 weights and 4 neurons. That will make a
fine VizANN. Do you know which of these ANN is the one?

I'll try and compile and run this soon; mebbe tomorrow.

m
--
I'm proud of http://ANNEvolve.sourceforge.net. If you want to write software, or articles, or do testing or research for ANNEvolve, let me know.



-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642