CiteSeerX — Table search results

Results 1 - 10 of 21,174

Table 2: Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c example. The sym- bol D indicates an arbitrary digit.

in Learning to Classify Text from Labeled and Unlabeled Documents

by Kamal Nigam, Andrew Mccallum, Sebastian Thrun, Tom Mitchell

"... In PAGE 5: ... Naturally, more unlabeled data helps. Table2 provides a window into the evolution of the classi er over the course of EM iterations for a speci c example. Based on the WebKB data set, each column shows the ordered list of words that the model believes are most predictive of the course class.... ..."

Table 2: epsilon1c of 4 arbitrary systems. The details are in [18].

in 1 PHENOMENOLOGY OF NEURAL AND SOCIAL NET- WORKS

by Marko Puljic, Robert Kozma 2004

"... In PAGE 7: ...alue is the same for all equivalent systems. The idea is explained in more detail in [16] and [17]. The epsilon1c values of many systems are presented in [18]. Table2 presents epsilon1c values of 4 arbitrary subsystems. epsilon1c is the point of state change.... ..."

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

in Text Classification from Labeled and Unlabeled Documents using EM

by Kamal Nigam, Andrew Kachites Mccallum, Sebastian Thrun, Tom Mitchell 2000

Cited by 490

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

in Text Classification from Labeled and Unlabeled Documents using EM

by Kamal Nigam , Andrew McCallum, Sebastian Thrun, Tom Mitchell 2000

Cited by 490

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

in Text Classification from Labeled and Unlabeled Documents using EM

by Kamal Nigam, Andrew Kachites McCallum, Sebastian Thrun, Tom Mitchell 2000

Cited by 490

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

in Text Classification from Labeled and Unlabeled Documents using EM

by Kamal Nigam, Andrew Mccallum, Sebastian Thrun, Tom Mitchell 2000

Cited by 490

Table 4. Estimated time (arbitrary units) by design phase for

in A Semi-custom Design Flow in High-performance Microprocessor Design

by Gregory A. Northrop, Pong-Fei Lu 2001

"... In PAGE 6: ... Instead, we made an effort to estimate the improvement, giving quantities that were derived from discussions with several designers. A proto-typical set of times are given in Table4 for 7 stages of design, for the full custom approach and the equivalent design in semi-custom. Once a design has been through the complete semi-custom flow once, an iteration typically only requires a couple of days, including structural changes to the design.... ..."

Cited by 3

Table 6 shows the result after we subsequently classified the modified dataset. The substantial dif- ference between results can be seen. On the other hand, if the record No. 7 is chosen as shown in Table 7, the result shown in Table 8 is also different from the original one. As it can be expected, the set of rules is substan- tially different from the original, even only one record is changed. Obviously, rules hiding by arbitrary mod- ification causes too much side effect.

in A Reconstruction-based Algorithm for Classification Rules Hiding

by Juggapong Natwichai, Xue Li, Maria E. Orlowska 2006

"... In PAGE 4: ... Therefore, unpruned classification rules, less significant rules, are used in the algorithm. Finally, the decision tree is used to generate a new Table6 : Modified credit card classification rules (The record No. 3 has been modified) Antecedence Class gender = female NO gender = male YES Table 7: Modified credit card approval dataset (The record No.... ..."

Cited by 4

Table 2. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c example. By the second iteration of EM, many common course-related word have high weights. The symbol D indicates an arbitrary digit.

in Learning to Classify Text from Labeled and Unlabeled Documents

by Kamal Nigam, Andrew Mccallum, Tom Mitchell 1998

"... In PAGE 12: ... To gain some intuition about why EM works, we present a detailed trace of one example. Table2 provides a window into the evolution of the classi er over the course of EM iterations for this example. Based on the WebKB data set, each col- umn shows the ordered list of words that the model believes are most \predictive quot; of the course class.... ..."

Cited by 139

Table 2.2: Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

in Using Unlabeled Data to Improve Text Classification

by Kamal Paul Nigam 2001

Cited by 37

Results 1 - 10 of 21,174

Tools

Table 2: Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c example. The sym- bol D indicates an arbitrary digit.

Table 2: epsilon1c of 4 arbitrary systems. The details are in [18].

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

Table 3. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.

Table 4. Estimated time (arbitrary units) by design phase for

Table 2. Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c example. By the second iteration of EM, many common course-related word have high weights. The symbol D indicates an arbitrary digit.

Table 2.2: Lists of the words most predictive of the course class in the WebKB data set, as they change over iterations of EM for a speci c trial. By the second iteration of EM, many common course-related words appear. The symbol D indicates an arbitrary digit.