Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
332,100
Table 1 Example of cyber-attack classifications in table format [20] Attack
"... In PAGE 12: ... We classify a set of example attacks which are being used in our research group to demonstrate one use of the SFR framework. From Table1 , we can extract common details among our subset of attacks to aid in our understanding. Then, as we discover characteristics of attacks in our ongoing research, we can compare those characteristics with those of similar attacks within the framework [22].... In PAGE 12: ... Then, as we discover characteristics of attacks in our ongoing research, we can compare those characteristics with those of similar attacks within the framework [22]. We describe two attacks in Table1 (UDP Storm and Slammer Worm) and how we choose the elements in the table for each attack. 5.... In PAGE 14: ... This allows us to meet the various needs of understanding and assessing the overall risks of attacks, and designing groups of protection means to reduce costs and improve efficiency. For example, in Table1 , eleven of the fifteen attacks classified originate from a remote location. Thus, a defense mechanism disallowing remote access to a network can defend against all of these attacks simultaneously.... ..."
Table 7: Example dialogue responses illustrating the differences between the DataDriven1 approach to content selection and response generation (at top) and the Bruiser system (at bottom).
2006
"... In PAGE 4: ... Although this provides a mechanism for de- scribing the backend data, content selection is still hand- crafted to the extent of defining thresholds and, perhaps more critically, associations between attributes are left un- explored. Table7 shows an example of output derived using the DataDriven1 system and Bruiser. There are two important distinctions between Bruiser and DataDriven1.... In PAGE 4: ... Although both systems may choose to speak about the same set of attributes, the Bruiser system, us- ing tree induction to discover association rules, allows us to speak about relationships among attributes, which gives the user a better understanding of the domain. An examination of the two dialogues shown in Table7 il- lustrates these differences. Both dialogues are driven by the same simulated user utterances.... In PAGE 4: ... Both dialogues are driven by the same simulated user utterances. The dialogue at the top of Table7 , the one using the DataDriven1 approach, has terser responses that contain less information about the backend data. The dialogue at the bottom of Table 7, using the Bruiser system, shows the result of learning association rules from the data (e.... ..."
Cited by 1
Table 1. Classification of sciences by motivation for international collaboration Data-driven
2004
"... In PAGE 4: ...6 Scientometrics 62 (2005) Table1 provides a notional concept of this typology, showing some fields of science as they would be classified within this scheme. Table 1.... ..."
Table 3. Results for the parallel combination of our 2 data-driven methods with a rule based system, using a Winner Take All (WTA) approach.
Table 1: Relative entropies for model-driven and data-driven FAN classifiers.
2004
"... In PAGE 7: ... Next to the occurence of such discrepancies, which can only be identified by having sufficient knowledge about the domain, the construction of an accurate classifier based on a small database is impaired in principle. The conjecture that suboptimal dependencies were added is supported by the increasing relative entropy between the declarative model and data-driven classifiers with increasing structural complexity ( Table1 ). It is unlikely that the naive classifier is simply the best representation of the dependencies within the model since relative entropy was shown to decrease for model-driven classifiers of increasing structural complexity.... ..."
Table 4: Comparison of baseline (A) and data-driven pronuncia- tion modeling method (B) on 96 Hub 4 DEV male data.
2000
Cited by 6
Table 1 shows the number of binary branching nodes for each of the two decision tree models for both English and German. The complexity of these decision trees validates the data-driven approach, and makes clear how daunting it would be to attempt to account for the facts of comma insertion in a declarative framework.
2002
"... In PAGE 4: ... Table1 Complexity of the decision tree models in Amalgam At generation time, a simple algorithm is used to decide where to insert punctuation marks. Pseudo-code for the algorithm is presented in Figure 1.... ..."
Cited by 1
Table 1: Results of decomposing/classifying the mixed pixels of image 2a. Thresholds were set at 24.0 (data-driven decomposition) and 848.0 (data-driven classi cation).
1996
"... In PAGE 15: ...0. The results of the experiments on image 2a are presented in Table1 . Data-driven decomposition needed one iteration in which the remaining 16 pixels were processed.... ..."
Cited by 3
Table 1: Results of four area estimation algorithms on images of di erent resolution. For data-driven decomposition the threshold is set to the standard value of 12, for data-driven classi cation it is set to an optimal value that di ers with the resolution (low for images of low resolution, high for high-resolution images).
"... In PAGE 13: ... As a result, it was expected that the error given in pixels would decrease, and the error expressed in mm2 would increase; the error given as a percentage of the total area, however, was unpredictable. The results of the experiment are presented in Table1 . For all area estimation algorithms it can indeed be observed that eA expressed in pixels decreased and eA expressed in mm2 increased as the IFOV got larger.... ..."
Results 1 - 10
of
332,100
