ProceedingsIEEE, 1997 - Artificial intelligence |
From inside the book
Results 1-3 of 39
Page 179
... weight is used to differentiate equivalent operations from basic operations . It is better to use low values . As the weight should not change too much the distance , we decide arbitrarily a weight of p , 2p , 3p , ... for equivalent ...
... weight is used to differentiate equivalent operations from basic operations . It is better to use low values . As the weight should not change too much the distance , we decide arbitrarily a weight of p , 2p , 3p , ... for equivalent ...
Page 240
... Weight change A. Weight storage Weight storage is a crucial element in a neural net re- alisation . Because of limitations in the circuit technology available for the time being in the LC chip the weights are stored in 4 - bit ...
... Weight change A. Weight storage Weight storage is a crucial element in a neural net re- alisation . Because of limitations in the circuit technology available for the time being in the LC chip the weights are stored in 4 - bit ...
Page 241
... weights on the chip and the num- ber of neural net I / O pins should be matched in such a way that for a given number of I / O pins the weight for all possible combinations of input and output are available . For the reconnection of the ...
... weights on the chip and the num- ber of neural net I / O pins should be matched in such a way that for a given number of I / O pins the weight for all possible combinations of input and output are available . For the reconnection of the ...
Contents
Innovations in Design and Integrated | 1 |
Creative and Innovative Manufacturing | 4 |
Y Fan and C Wu Tsinghua Univ | 7 |
Copyright | |
43 other sections not shown
Other editions - View all
Common terms and phrases
according activities algorithm allows analysis application approach architecture assembly ATRT Automation called communication complete components considered constraints control system cost customers decision defined described determined dispatching distributed dynamic effective Engineering enterprise environment error estimated example exists Figure fixture flexible flow function given implementation important improvement industrial input integration interaction interface INTRODUCTION knowledge lead learning load locating machine manuals manufacturing means measurements method nets object obtained operation optimal output parameters performance Petri planning plant position possible presented problem proposed prototype real-time reference represents requirements robot rules scheduling selected sequence server shown shows simulation solution specific standard station structure task token tool transition unit University virtual