"... In PAGE 2: ... This can be used to find the fundamental differences between new problem definitions and heuristics and previous ones. If the problem definition already exists, the references in Table 1 (or Table2 in the next Section) can be consulted to find suitable heuristics for the problem. To limit the scope of this paper, we do not consider al- gorithms that aim solely at improving the availability or reliability of the system, nor algorithms dealing only with migration.... In PAGE 5: ...roblem definition. E.g., the problem definition might spec- ify the cost function C8CXBEBV C8CYBEC6 D6CTCPCSD7CXCZ A1 CSCXD7D8CXCY A1 DDCXCYCZ, but the Hotspot heuristics disregards the distance and uses just C8CXBEBV C8CYBEC6 D6CTCPCSD7CXCZ A1 DDCXCYCZ. Table2 lists previously proposed heuristics for replica placement problems, mapped into our framework. The ta- ble serves mainly two purposes.... In PAGE 5: ... Note that the table and the primitives only give a high-level view of the heuristics, and in order to faithfully implement one the source reference has to be consulted. For example, while the tree-based heuristics are grouped together in Table2 , their respective implementations for the six problem definitions are different. Heuristics from [28, 29, 31, 32] are not included due to space constraints.... In PAGE 5: ... The link metric scope is not shown in the table as none of the heuristics use it. Table2 indirectly reflects if a heuristic is an off-line or an on-line approach. An off-line heuristic would start from scratch every time it is executed and make the placement decision for all objects and nodes it considers.... In PAGE 8: ...2 Goodness of Existing Algorithms Figure 3 illustrates the cumulative distribution function of the client-perceived latency for a system with 1000 nodes, 100 objects and 32 replicas of each object. We focus on three problem definitions used in CDNs (total distance A1 demand, total distance and max distance) and a number of previously proposed heuristics taken from Table2 . There are three main points to take away from the graphs.... ..."

"... In PAGE 6: ....3. Knowledge Resources Three of the frameworks explicitly address the knowledge resources dimension by identifying different kinds of knowledge resources. These are summarized in Table2 . Leonard-Barton [7] identifies two types of knowledge resources: employee knowledge and physical systems (e.... In PAGE 7: ... The frameworks described and compared here can serve as a starting point for creating a generic framework that unifies KM concepts. Such a framework should be sufficiently comprehensive to address all the main content features of those presented in Table2 , 3 and 4. It should also accommodate other concepts appearing in the KM literature, but outside the scope of extant KM frameworks.... ..."

"... In PAGE 9: ... By restricting the feature space from which the predictive attributes are selected to collaborative features (attributes derived only from the Order table), content features (attributes derived from the Order, Book, and Occurrence tables) and demographic features (attributes derived from the Order and Customer tables) we also built models that emulate collaborative filtering, demographic filtering, content-based approaches. The performances of different recommendation approaches under this PRM-based recommendation framework (PRMR) are presented in Table2 in comparison with the performance of a standard collaborative filtering algorithm based on customer neighborhood formation (Breese et al. 1998).... In PAGE 10: ... Book Store dataset: Recommendation performance measures (K=10) of a standard collaborative filtering algorithm and the different models of a unified relational-learning- based recommendation framework (PRMR) that emulate various typical recommendation approaches (boldfaced measures were not significantly different from the largest measure at the 5% significance level). We observe in Table2 that the content-based and demographic filtering approaches had similar performances, the performance measures of the collaborative filtering approach almost doubled those of the content-based and demographic filtering approaches, and that the hybrid approach delivered the best performance with significant improvement compared to the collaborative filtering approach. All PRM-based models under different approaches outperformed the standard neighborhood-based collaborative filtering algorithm, which demonstrates the value of additional recommendation-relevant relational features constructed by the PRM-based recommendation framework that are not included in typical recommendation algorithms.... ..."

"... In PAGE 9: ... The Table 3, row 2 shows the number of correctly identified samples. As shown in the Table3 , the DEGs obtained by the unified framework offered 100% accuracy in the identification of training sample classes. The two-way framework also offered 100% accuracy in identification of the labels of training Page 9 of 21 (page number not for citation purposes) Leukemia (ALL) and Acute Myeloid Leukemia (AML).... In PAGE 11: ...ttp://www.biomedcentral.com/1471-2105/8/347 The ASI algorithm is further applied to cluster the test samples using the DEGs obtained through respective methods. It is evident from the row 3 of Table3 that the DEGs obtained using the unified framework classified the AML and ALL samples better (97.06%) than the DEGs obtained using the other methods.... In PAGE 11: ...ithm (3DSCP). Comparing the Figs. 5(a) to 5(f) it may be seen that the unified framework offered clear differentia- tion between different sample cases. Although the two- way clustering and unified approach identified all the samples correctly as shown in the Table3 for training samples, comparing the Figs. 5(e) and 5(f), it may be seen that the unified framework offered much clear separation between the samples of different cases.... ..."

"... In PAGE 11: ... Gene selection and statistical validation The gene selection is performed such that there is mini- mum percentage of expected false positives. As shown in the Table4 , the unified framework recorded less percent- age of false positives (2.48%-11.... ..."