<HTML><FONT FACE=arial,helvetica><HTML><FONT SIZE=2 PTSIZE=10 FAMILY="SANSSERIF" FACE="Arial" LANG="0">I am using tlearn as bigXOR in doing mathematical encoding of faces as in visnet. The backpropogation means that the information I dont know in the hidden layers allows the model to learn as 0 or 1 and change the x or to make the output lesion and learn. <BR>
<BR>
<BR>
I use is to geology mapping, facial mapping and neuroscience as well as forensics as who fits the pattern adding the units I know and dont know. It is simple to learn and outclassed theories as it predicted that which I dismissed as improbably. Hence the Neurotool and freeware in connectionist book of McCleod from the Oxford University disc and examples in the book lets you share the way you use it and change the model to suit the need as free software. You will see how it is used in papers in future other than neuroscience as Codis is to dna lesioned thinking of machines.<BR>
<BR>
</FONT><FONT COLOR="#8000ff" BACK="#ffffff" style="BACKGROUND-COLOR: #ffffff" SIZE=4 PTSIZE=14 FAMILY="SCRIPT" FACE="CloisterBlack BT" LANG="0"><I>Paula 'humanity is intelligent thinking' beyond self'</I></FONT></HTML>