### Table 13: The operators of the computer domain.

1994

"... In PAGE 51: ... Similarly, for it to be plugged in the device must be located with the reachofapower outlet. Table13 shows the operators for this domain. For this domain, Alpine generates the graph shown in Figure 10.... ..."

Cited by 5

### Table 1: Number of adjustable weights required for single-output PSNs and HPUs of order K, when the input is N-dimensional. network to be regular and incrementally expandable, since the order can be increased by adding another summing unit and associated weights, but without disturbing any connection established previously. The number of weights for a PSN is linear in input size and much smaller than that of an HPU. For the purpose of comparison, we summarize in Table 1 the required number of adjustable weights for PSN and HPU with single output when input dimension is N and all the higher-order terms of up to K-th order are to be incorporated. PSNs can handle both analog and binary input/output by using a suitable nonlinear activation function ( ). The analog pi-sigma network (APSN) is a PSN that uses the sigmoid activation function:

1992

"... In PAGE 30: ...3 69.8 Table1 0: Comparison of APSN and BPSN for N bit parity check problem, K = N. the negation problem requires multiple outputs.... In PAGE 32: ...9 16.7 Table1 1: The simulation results for N bit symmetry and negation problem using BPSN with the asynchronous updating rule. depending on whether the most signi cant bit of the input is 1 or not.... In PAGE 32: ... (i) Parity problem: APSN = 1:0; HPU = 0:2; MLP = 0:2, (ii) symmetry problem: APSN = 0:2; HPU = 0:2; MLP = 0:1, and (iii) negation problem: APSN = 1:0; HPU = 1:0; MLP = 0:25. As can be observed in Table1 2, PSN yields comparable or better results than HPU, and both networks perform much better (i.... In PAGE 33: ...6 34.3 9120 Table1 2: Comparison of the APSN, HPU and MLP for Boolean function realization. For all problems, APSN and HPU use 4th degree polynomials.... ..."

Cited by 4

### Table 1 Process disturbances

in Comparison of statistical process monitoring methods: application to the Eastman challenge problem

### TABLE III Disturbance

### TABLE IV Disturbance

### Table 4: Errors for load disturbances

2005

"... In PAGE 89: ... The initial feedforward signals equal the feedforward signals obtained without disturbances. The maximum calculated absolute errors and the max- imum absolute errors obtained in the simulations show a good correspondence as can be seen in Table4 . The presence of the load disturbances and model uncertain- ties in the learning process leads to an amplification of... ..."

### Table 3. Impact of the added disturbance.

2006

"... In PAGE 16: ... Figure 7 shows the generated paths corresponding to four increasing values on the noise parameter. As can be seen in the third column of Table3 , the disturbance corresponds to a displacement of more than 40%. Table 3 offers additional insight into the simulations made.... In PAGE 16: ... As can be seen in the third column of Table 3, the disturbance corresponds to a displacement of more than 40%. Table3 offers additional insight into the simulations made. In addition to the particular values of w that 16 of 25... ..."

Cited by 2