Results 1 - 10 of 268

Table 6.6: Substitution matrix and pseudocounts for regularizer using substitution matrix plus scaled counts plus pseudocounts (trained on jsj = 0; 1; 2; 3). The feature alphabets have very few tuning parameters (one per alphabet), and so one would expect them not to do well relative to the pseudocount methods (20 parameters) or the substitution matrices (400 parameters). The excess entropy reported in Table 6.7 show the feature alphabets doing surprisingly well for having so few parameters. The 8-alphabet set does quite well for the the samples sizes it was tuned for (jsj = 1; 2), but degrades rather rapidly for larger sample sizes, doing worse than zero-o sets by jsj = 4. The 4- alphabet and 5-alphabet sets do better than pseudocounts for jsj = 1; 2; 3, but, like the substitution matrix method, the feature alphabets continue to get further from the optimum regularizer as jsj increases, while the pseudocount methods improve. The results for hand-created feature alphabet sets is presented in Table 6.8. On the whole, these hand-created feature sets did not do as well as the automatically generated ones. Although the tiny number of parameters for the feature alphabets makes them aesthetically appealing, their performance is not good enough to justify the e ort of implementing them. Perhaps

in Regularizers for Estimating Distributions of Amino Acids from Small Samples
by Kevin Karplus 1995
Cited by 21

Table 2: The specification of target constructs for the questionnaire development

in unknown title
by unknown authors
"... In PAGE 3: ...se (1998, p. 2). This definition was given to all the participants in this study to clarify what usability means. Based on the descriptive definition, aesthetic appeal (image or design) (Ketola, 2002; Kwahk, 1999) and emotional dimensions (Jordan, 2000; Logan, 1994) were added as important sub-dimensions, since the target products are consumer products not software products. The summary for the conceptualization of the target construct is provided in Table2 . This specification was referred to as the target construct throughout this study.... ..."

Table 3: Prescriptive Aesthetic Framework Based on Data and Aesthetic Literature

in THEORY AND MODELS FOR CREATING ENGAGING AND
by Immerslve Ecommerce Websltes, Morgan Jennings

Table 9: Structure Appeal Questionnaire

in unknown title
by unknown authors 2004
"... In PAGE 47: ... Eight dimensions of this questionnaire were translated and used. Table9 shows the structure of the questionnaire. A Liker scale (1-7) was used.... ..."

Table 2) for predicting choices of aesthetically pleasing

in THEORY AND MODELS FOR CREATING ENGAGING AND
by Immerslve Ecommerce Websltes, Morgan Jennings

Table 1 shows the aesthetic values for all the diagrams.

in Graph drawing aesthetics and the comprehension of UML class diagrams: an empirical study
by Helen C. Purchase, Matthew Mcgill, Linda Colpoys, David Carrington

Table 5. Aesthetic criteria for sequence diagrams.

in How to Draw a Sequence Diagram
by Timo Poranen, Erkki Mäkinen, Jyrki Nummenmaa

Table 6. Computational complexity of the aesthetics for sequence diagrams.

in How to Draw a Sequence Diagram
by Timo Poranen, Erkki Mäkinen, Jyrki Nummenmaa

Table. An overview of the results of assessment by aesthetic judgement.

in An aesthetic comparison of rule-based and genetic algorithms for generating melodies
by Andrew R. Brown

TABLE 2. Global appealing of the user interface.

in Individual Differences in a Spatial-Semantic Virtual Environment
by Chaomei Chen 2000
Cited by 16
Results 1 - 10 of 268