### Table 3: Serbo-Croatian Recognition Results based on Adapted Vocabulary using Morphological Knowledge.

1997

Cited by 7

### TABLE V CLASSIFICATION ACCURACY USING RBFNN MODELS THAT INCORPORATE PRIOR KNOWLEDGE BASED ON AMINOACID CHEMICAL PROPERTIES (DECIMAL ENCODING) AND DIMENSIONALITY REDUCTION BASED ON MOTIF DISCOVERY Classification Accuracy

### Table 2. Eye fixations using browsers.

"... In PAGE 8: ... Fixations. As shown in Table2 , on average participants using the Hyperbolic browser made more fixations per task than with the Explorer browser [F(1, 787) = 16.67, MSE = 0.... ..."

### Table 1: Serbo-Croatian Recognition Results based on Adapted Vocabulary using Morphological Knowledge and Phonetic Dis- tances.

"... In PAGE 4: ...5% 7.1% Table1 0: Results on different methods to create the Fallback Lex- icon. 8.... ..."

### Table 1: Component categories and their facets.

"... In PAGE 40: ...b1 b2 b3 b4 b5 2 6 6 6 4 3 7 7 7 5 t1 915 765 500 525 628 t2 500 1195 500 500 524 t3 500 529 500 500 500 t4 530 500 548 1049 669 t5 500 500 500 629 952 (27) D(f) = b1 b2 b3 b4 b5 2 6 6 6 4 3 7 7 7 5 t1 0:410 0:536 0:004 0:613 0:448 t2 0:413 0:807 0:609 0:002 0:441 t3 0:597 0:637 0:447 0:442 0:002 t4 0:704 0:005 0:535 0:808 0:640 t5 0:002 0:703 0:406 0:412 0:596 (28) nal scheme target base base t1 b3 b3 t2 b4 b4 t3 b5 b5 t4 b2 b2 t5 b1 b1 Table1 0: Results of the nal mapping of two isomorphic graphs with 5 vertices. 4.... In PAGE 42: ...denotes whether a one-to-one or many-to-one mapping exist between the target and base graph vertices. (1) num ber of correct mappings (%) (2) one-to-one mapping? 5 vertices 100 yes 6 vertices 100 yes 7 vertices 57 no 9 vertices 78 no library/wholesaler 78 yes (9 vertices) Table1 1: Summary of the mapping performances. An important issue is how much computation the proposed mapping algorithm requires in order to reach a solution.... In PAGE 43: ... There is no clear tendency that the number of epochs for the LRAAM network increases when the number of vertices in the graphs increases. Neither the dimension of the reduced descriptions (column 6) nor the maximum number of branches for the graphs (column 7 in Table1 2) seem to have a clear e ect on the total number of epochs for the LRAAM networks. Even though the mapping algorithm in the 9 vertices case and library/wholesaler case cycled for 10 iterations, the LRAAM network for the base graph in the rst case was not invoked more than 3 times, whereas 9 times in the latter case.... In PAGE 44: ...iterations of mapping algorithm (2) num ber of LRAAM base graph learnings (3) total epochs of LRAAM net works (4) lowest num ber of LRAAM epochs (5) highest num ber of LRAAM epochs (6) dimension of reduced descriptions (m) (7) maxim um branc hing (n) 5 vertices 10 3 1745 143 1120 3 2 6 vertices 3 2 780 211 294 3 4 7 vertices 10 8 7492 331 1509 3 2 9 vertices 10 4 1149 149 308 3 6 library/wholesaler 10 9 11363 864 1411 4 6 (9 vertices) Table1 2: Computational measurements of the mapping algorithm for the experi- ments in Section 4. 5 vertices 436 6 vertices 260 7 vertices 832 9 vertices 230 library/wholecase 1136 (9 vertices) Table 13: Average number of learning epochs for the LRAAM networks used in the experiments in Section 4.... In PAGE 44: ...iterations of mapping algorithm (2) num ber of LRAAM base graph learnings (3) total epochs of LRAAM net works (4) lowest num ber of LRAAM epochs (5) highest num ber of LRAAM epochs (6) dimension of reduced descriptions (m) (7) maxim um branc hing (n) 5 vertices 10 3 1745 143 1120 3 2 6 vertices 3 2 780 211 294 3 4 7 vertices 10 8 7492 331 1509 3 2 9 vertices 10 4 1149 149 308 3 6 library/wholesaler 10 9 11363 864 1411 4 6 (9 vertices) Table 12: Computational measurements of the mapping algorithm for the experi- ments in Section 4. 5 vertices 436 6 vertices 260 7 vertices 832 9 vertices 230 library/wholecase 1136 (9 vertices) Table1 3: Average number of learning epochs for the LRAAM networks used in the experiments in Section 4.... ..."

### Table 2: Adaptive middleware categorized by middleware layers and adaptation type.

2003

"... In PAGE 31: ... Middleware Application Domain Layers QoS-Enabled Dependable Embedded Domain-Services Common- QuO OGS, IRL, FTS, Services TAO-LB, Racks TAO, CIAO, Orbix, OpenORB, FRIENDS, UIC, Distribution Squirrel, DynamicTAO, Electra, ZEN, FlexiNET, OpenCorba, AspectIX Orbix+Isis Orbix/E Host-Infra. ACE, Eternal, Isis, Horus, PersonalJava MetaSockets Ensemble, Totem, Rocks EmbeddedJava Table2 categorizes the adaptive middleware projects using the middleware layers and adaptation type. This table shows that adaptive middleware research has exploited both static and dynamic adaptations.... In PAGE 31: ... Safe adaptation can harness the dangerous power of dynamic adaptation provided by mutable middleware. Table2 also shows that... ..."

Cited by 7

### Table 1: Results in Discovering People and Associated Home Page. The flrst line refers to accuracy in discovering names using the procedure for seed names (Citeseer+Google, etc.), the second one to the discovery using adaptive IE.

### Table 2: Adaptive middleware categorized by middleware layers and adaptation type.

"... In PAGE 31: ... Middleware Application Domain Layers QoS-Enabled Dependable Embedded Domain-Services Common- QuO OGS, IRL, FTS, Services TAO-LB, Racks TAO, CIAO, Orbix, OpenORB, FRIENDS, UIC, Distribution Squirrel, DynamicTAO, Electra, ZEN, FlexiNET, OpenCorba, AspectIX Orbix+Isis Orbix/E Host-Infra. ACE, Eternal, Isis, Horus, PersonalJava MetaSockets Ensemble, Totem, Rocks EmbeddedJava Table2 categorizes the adaptive middleware projects using the middleware layers and adaptation type. This table shows that adaptive middleware research has exploited both static and dynamic adaptations.... In PAGE 31: ... Safe adaptation can harness the dangerous power of dynamic adaptation provided by mutable middleware. Table2 also shows that... ..."

### Table 5. Inclusion or Non-inclusion of Facets of Our Model in Previous Models

"... In PAGE 16: ... Table 5 lists topics in our model and indicates whether those topics are or are not included explicitly in the models in Table 2. Table5 is not meant as a criticism of the previous models because each of those models was designed for a particular purpose and served that purpose. In contrast to the previous models, our risk model is more complete, adaptable, and practical for use by business managers.... ..."

Cited by 2