CiteSeerX — Table search results — generated false alarm

Results 1 - 10 of 31,964

Table 3. Effects of filtering the training data on the performance in test phase Application Number of attacks detected Number of attacks detected (false alarms generated) - stide (false alarms generated) - LERAD Without filtering With filtering Without filtering With filtering

in Data Cleaning and Enriched Representations for Anomaly Detection in System Calls

by Gaurav T, Philip Chan, Debasis Mitra

Table 2: False Alarms Generated by RALPH-WS for Truck Drivers with no Virtual Lane Boundary

by Thesis Proposal Parag, Parag H. Batavia

"... In PAGE 21: ... This is not entirely true, however, as RALPH does allow the user to exceed the lane boundary by a tunable amount. Table2 gives the numbers of alarms triggered by motion to the left and the right. These alarms are further broken down by roadway geometry, i.... ..."

Table 3: Detection and False Alarm Probability of Linear Prediction

in Extension And Alarm Analysis Of

by Neural Network Prediction, Nattavude Thirathon

"... In PAGE 5: ...able 2: Neural Network Prediction Error.......................................................................... 8 Table3 : Detection and False Alarm Probability of Linear Prediction .... In PAGE 14: ...75 Almost all of the alarms generated by the linear prediction are false alarms. Neural network is promising in reducing the false alarm rate as shown in Table3 . Neural network eliminates many false alarms.... ..."

Table 6. The 100 top false alarms detected by PHAD and ALAD.

in Learning Nonstationary Models of Normal Network Traffic for Detecting Novel Attacks

by Matthew V. Mahoney, Philip K. Chan

"... In PAGE 5: ... Duplicate detections of the same attack are counted only once, but every false alarm is counted. The top results reported by [9, Table6 ] at a false alarm rate of 10 per day (100 total) are shown in Table 3. Table 3.... In PAGE 8: ...2. Analysis of False Alarms Table6 shows the causes of the 100 false alarms in the evaluation. A few alarms were generated by more than one anomaly, so the total is more than 100.... ..."

Table 4.2: Fraction of distance estimated and associated standard deviation Figure 4.22 to Figure 4.25 show the performance with di erent epsilon ranges lowresTW . In Figure 4.22, the number of false alarms and false dismissals are shown for querying a database of sequence length of 256. We observe that when we scale the value of the epsilon only by a factor of dist frac with zero s:d:, i.e. lowresTW = 1:21 timewarp, there are 85 false dismissals 7 being generated. On 7Number of false alarms and dismissals should be compared to the average number of quali ed sequences in the answer set, which is = database size average epsilon timewarp = 5000 2.75 = 137.5.

in By CHAN, Kin Pong

by Prof Wai-chee, Fu Ada 1999

Table 7: Comparison of false alarm rates

in Evolutionary Design of Intrusion Detection Programs

by Ajith Abraham, Crina Grosan, Carlos Martin-vide 2006

"... In PAGE 13: ... The values of TPR and FPR obtained by MEP and GEP for the test data set are depicted in Table 7. The true positive rate (TP) is given by: TP = positives correctly classifled total positives The false positive rate (FP) is given by: FP = total negatives - negatives incorrectly classifled total negatives As evident from Table7 , when compared to LGP, MEP obtained the best values for TPR and FPR for Normal and DoS and good performance for the other classes. MEP required only 30 and 20 generations for the Normal and U2R types respectively.... ..."

Cited by 2

Table 19: Effect of Advert Detection on diarisation score, when using RT-03s diarisation clustering on the bndidev03 and bneval03 data. Bracketed numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

in An Investigation into the Interactions between

by Speaker Diarisation Systems

"... In PAGE 32: ...2.1 Effect of Removing Adverts on Diarisation Performance The effect of the advert detection when using the CUED RT-03s diarisation system clustering is given in Table19 for the bndidev03 and bneval03 data. Results with no advert detection, and perfect advert detection33 are also given for comparison.... ..."

Table 20: Effect of Advert Detection on diarisation score and WER, when using RT-03s STT clustering on the bndidev03 data. Bracketed diarisation numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

in An Investigation into the Interactions between

by Speaker Diarisation Systems

"... In PAGE 33: ... 2003b). The diarisation scores are given in Table20 along with the WER from running the CUED 10xRT BN STT system described in section 3.2.... ..."

Table 19: Effect of Advert Detection on diarisation score, when using RT-03s diarisation clustering on the bndidev03 and bneval03 data. Bracketed numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

in An Investigation into the Interactions . . .

by S. E. Tranter, K. Yu, D. A. Reynolds, G. Evermann, D. Y. Kim, P. C. Woodland 2003

"... In PAGE 33: ...2.1 Effect of Removing Adverts on Diarisation Performance The effect of the advert detection when using the CUED RT-03s diarisation system clustering is given in Table19 for the bndidev03 and bneval03 data. Results with no advert detection, and perfect advert detection33 are also given for comparison.... ..."

Table 20: Effect of Advert Detection on diarisation score and WER, when using RT-03s STT clustering on the bndidev03 data. Bracketed diarisation numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

in An Investigation into the Interactions . . .

by S. E. Tranter, K. Yu, D. A. Reynolds, G. Evermann, D. Y. Kim, P. C. Woodland 2003

"... In PAGE 34: ... 2003b). The diarisation scores are given in Table20 along with the WER from running the CUED 10xRT BN STT system described in section 3.2.... ..."

Results 1 - 10 of 31,964

Tools

Table 3. Effects of filtering the training data on the performance in test phase Application Number of attacks detected Number of attacks detected (false alarms generated) - stide (false alarms generated) - LERAD Without filtering With filtering Without filtering With filtering

Table 2: False Alarms Generated by RALPH-WS for Truck Drivers with no Virtual Lane Boundary

Table 3: Detection and False Alarm Probability of Linear Prediction

Table 6. The 100 top false alarms detected by PHAD and ALAD.

Table 7: Comparison of false alarm rates

Table 19: Effect of Advert Detection on diarisation score, when using RT-03s diarisation clustering on the bndidev03 and bneval03 data. Bracketed numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

Table 20: Effect of Advert Detection on diarisation score and WER, when using RT-03s STT clustering on the bndidev03 data. Bracketed diarisation numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

Table 19: Effect of Advert Detection on diarisation score, when using RT-03s diarisation clustering on the bndidev03 and bneval03 data. Bracketed numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.

Table 20: Effect of Advert Detection on diarisation score and WER, when using RT-03s STT clustering on the bndidev03 data. Bracketed diarisation numbers are generated not excluding any regions from scoring, and thus treat adverts and vocal noise as false alarm regions if hypothesised as speech.