### Table 5: Model architectures used in coarse-grained map prediction. NHO = number of hidden units in output network of the 2D RNN. NHH = number of hidden units in four hidden networks of the 2D RNN. NOH = number of output units in four hidden networks of the 2D RNN. NENC, NHEO = number of output, hidden units of the output network of the encoding BRNN. NHEH, NOEH = number of hidden, output units of the network associated with the chains of the encoding BRNN.

2003

"... In PAGE 14: ... For example in the case of NHH = NOH = NHO = NHEH = NOEH = NHEO = 8, NENC = 6 and for |Ie| = 24 (20 for amino acid frequencies, 3 for secondary structures, and 1 for solvent accessibility) the total number of parameters in the BRNN is 3,568. The total number of parameters for all the models used in the simulations is reported in Table5 . Learning rates and algorithms are similar to to the case of fine-grained maps.... ..."

Cited by 22

### Table 5: Model architectures used in coarse-grained map prediction. NHO =numberof hidden units in output network of the 2D RNN. NHH = number of hidden units in four hidden networks of the 2D RNN. NOH = number of output units in four hidden networks of the 2D RNN. NENC, NHEO = number of output, hidden units of the output network of the encoding BRNN. NHEH, NOEH =number of hidden, output units of the network associated with the chains of the encoding BRNN.

2003

Cited by 22

### Table 13: Iso-value Contour Surface speci- fications for the algorithm used to generate the image D of Figure 11.

2001

"... In PAGE 28: ...around surfaces (see Table13 in the Appendix). The differences can be measured and analyzed similar to the top row in Figure 11 using both image and data level approaches.... In PAGE 36: ... We also give a function table that maps the given gradient magnitude to a value to be multiplied to the color(or opacity) values resulted from the classification function [13]. Table 10 to Table13 provide the surface classification definitions for the algorithm used in Figure 11 (A - D). We gave the data range and color intensity for Binary Surface classification.... ..."

### Table 2: Contour selection results for House image.

1997

"... In PAGE 9: ... 5 a, b, c, and d, respectively. The quantitative results for this test are shown in Table2 and 3. The qualitative and quantitative evaluation of results indicates that the selection scheme works quite successfully.... ..."

Cited by 1

### Table 3: Contour selection results for Lenna image.

1997

Cited by 1

### Table 1: Hypothetical Two-Way Two-Mode Binary Data and Hierarchical Classes Model

2003

"... In PAGE 5: ...5 a hierarchical classes model (of a prespecified complexity) exists. Throughout this section, we will use the hypothetical child by item matrix in the top panel of Table1 as an example of a reconstructed data matrix. The items are sums of fractions presented to the children, and we denote yij = 1 if child i succeeds in item j and yij = 0 otherwise.... In PAGE 5: ... The association relation is the binary relation between the row elements and column elements as defined by the 1-entries in hatwide Y . From the data matrix in Table1 , for example, we can read that parenleftbigJohn, 5 9 + 7 9 parenrightbig is an element of the relation between children and items, while parenleftbigJohn, 2 3 + 1 2 parenrightbig is not. Two equivalence relations are defined, one on the row elements and one on the column elements of hatwide Y .... In PAGE 6: ... In particular, the example was constructed with three underlying abilities in mind: (I) finding the lowest common multiple of two integers, (II) finding the greatest common divisor of two integers, and (III) dealing with fractions where the numerator exceeds the denominator. The bottom panel of Table1... In PAGE 7: ... (1995) proposed a graphical representation that gives a full account of the relations represented by the conjunctive hiclas model. A graphical representation of the model in Table1 is found in Figure 1. Child and item classes appear as paired boxes, the upper box of each pair being a child class and the lower box an item class.... ..."

Cited by 1

### Table 3: Binary representation of nucleotides.

2007

"... In PAGE 8: ... Finally, the algorithm is extended to achieve an even higher speed-up. Binary coding To fit 64 nucleotides into 128 bits each nucleotide is rep- resented by two bits as described in Table3 . This coding has been chosen to simplify the comparison of nucle- otides.... ..."

### Table 1: Melodic contour intervals defined for 5 step representation.

2004

Cited by 4