Results 1 - 10 of 45,320

Table 7 Achievable gains of two externally informed methods using a real and a perfect FP detector, respectively

in Coping with disfluencies in spontaneous speech recognition: Acoustic . . .
by Frederik Stouten , Jacques Duchateau , Jean-Pierre Martens , Patrick Wambacq
"... In PAGE 15: ...7 and the average number of corrected words per FP, denoted as #cwrds/FP. Our results ( Table7 ) do confirm that with a per- fect FP detector, the WER can be reduced by 4.1% (0.... ..."

Table 3. Most relevant features for each clas label and information gain measures

in Abstract
by H. Güneş Kayacık, A. Nur Zincir-heywood, Malcolm I. Heywood
"... In PAGE 3: ...lases N is set to 10, (i.e. maximum 50,00 partitions). 3 Results Results are presented in terms of the clases that achieved god levels of discrimination from others in the training set and the analysis of feature relevancy in the training set. Table3 details the most relevant features for each clas and provides the coresponding information gain measures. Thre clases (namely normal, neptune and smurf) stand out from others with high information gain, hence high degre of discrimination.... ..."

Table 3: Deceptive Information Gain

in Greedy Attribute Selection
by Rich Caruana, Dayne Freitag 1994
"... In PAGE 5: ...5 This is not, however, the case in general. Consider a learning problem with three attributes, A, B, and C, that have the information gains and generalization performances shown in Table3 for ID3 grown trees. FS first considers the attributes in isolation.... ..."
Cited by 173

Table 7: Information gains g

in unknown title
by unknown authors 1999

TABLE II INFORMATION GAIN OF ATTRIBUTES

in Constructing Traffic-Aware Overlay Topologies: A Machine Learning Approach
by Benjamin D. Mcbride

Table 1: Performance gain achieved by S3

in Simultaneous Speculation Scheduling - A Technique for Speculative Dual Path Execution
by A. Unger, E. Zehendner, Th. Ungerer
"... In PAGE 6: ... Both processor models do not include any assumptions on caching ef- fects, because of the size of the used sample code. The values presented in Table1 were derived from counting the machine cycles, processors that match our processor models would need to execute the gen- erated programs. The numbers shown are the number of instructions in the program code, the average num- ber of instructions executed (instructions per thread, weighted with the measured execution probability of the thread), the number of conditional branches, the number of speculations performed by S3, the nest of speculation, the cycles for the SBP, the cycles for the MBP, and the performance gain (calculated as (SBP cycles / MBP cycles) -1).... ..."

Table 1: Performance gain achieved by S3

in Compiler Supported Speculative Execution on SMT Processors
by A. Unger, E. Zehendner, Th. Ungerer
"... In PAGE 7: ...ync. Here we expect these instructions to execute within a single cycle. Both processor models do not include any assumptions on caching e ects, because of the size of the used sample code. The values presented in Table1 were derived from counting the machine cycles, processors that match our processor models would need to execute the generated programs. The numbers shown are the number of instructions in the program code, the average number of instructions executed (instructions per thread, weighted with the measured execution probability of the thread), the number... ..."

Table 1 Direct and Indirect benefits gained by using the Internet among participants.

in Discussion Discussion
by unknown authors
"... In PAGE 3: ... (1993) have also pointed out that to achieve longer-term benefits from tel- ecommunications-based information sys- tems, an organisation needs to combine its inter-organisational systems strategy with existing business strategy. From the results of this study and also from our earlier survey (Poon and Swatman 1996), we observed that benefits gained from Internet access nominated by small businesses can also be classified into di- rect and indirect benefits, as Table1 illus- trates. An interesting fact is that firms A, D and E did not explicitly state that they had gained direct benefits from the Internet.... ..."

Table 4. Information gains of summary metrics

in A Semi-automatic Framework for Mining ERP Patterns
by Jiawei Rong, Dejing Dou, Robert Frank 2007
"... In PAGE 5: ... The metrics that is used to evaluate this dif- ferentiability is called information gain. Table4 shows the information gain of each attribute. In the rules provided by domain experts, the temporal criterion TI-max and the spa- tial criterion IN-mean (ROI) are used.... ..."
Cited by 1

Table 4. Average Information Gain of each attribute

in RECENT ADVANCES IN MECHANICS AND RELATED FIELDS UNIVERSITY OF PATRAS 2003 in Honour of Professor Constantine L. Goudas EFFICIENCY OF MACHINE LEARNING TECHNIQUES IN PREDICTING STUDENTS ’ PERFORMANCE IN DISTANCE LEARNING SYSTEMS
by S. B. Kotsiantis, C. J. Pierrakeas, I. D. Zaharakis, P. E. Pintelas
"... In PAGE 7: ... 303 The average information gain and the variance of each attribute according to the training data and the ten groups of data collected from 10 tutors for the academic year (2001-2) are presented in Table4 . The larger the value of the information gain is, the more influence of the attribute in the induction.... ..."
Results 1 - 10 of 45,320