Now 5233 visitors
Today:0 Yesterday:0
Total: 147758
212S 50P 4R
2026-10-20, Week 43 |
| Member Login |
| TACT Journal Homepage |
| Call for Paper |
| Committee |
| Find My Paper |
| Author Homepage |
| Statistics & History |
| Welcome Message |
| Paper Procedure |
| FAQ |
| Paper Archives |
| Outstanding Papers |
| Author Homepage |
| - Paper Procedure |
| - Journal Procedure |
| - Presentation Tips |
|
| Program & Proceedings |
| Program with Papers |
| Plenary Session |
| Tutorial Session |
|
| Presentation Platform |
| Hotel & Travel Info |
| About Korea |
| Accommodation |
| Transportation |
| VISA |
| Other Infomation |
|
| Photo Gallery |
| Scheduler Login |
| Seminar |
| Archives Login |
|
| Sponsors |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Save Q&A |
 |
* Edit or answer any Q&A by selecting Q&A number Hyper Link below + Write button (Save)
|
ICACT20260103 Question.1
Questioner: hripsime.hovhannisyan@rau.am 2026-02-10 오후 5:20:29 |
ICACT20260103 Answer.1
Answer by Auhor arjunvirvashisht@gmail.com 2026-02-10 오후 5:20:29
Chrome  Click!! |
| Regarding the practical application of this framework, what is the total parameter count? |
This is a critical point for practical cybersecurity. In our experimental setup, the average inference latency per URL is approximately 0.15 to 0.25 milliseconds on standard hardware. This efficiency is achieved because the final classification is offloaded to a Support Vector Machine (SVM) using the deep temporal embeddings extracted by the BiLSTM. For real-time line rate defense, our framework supports asynchronous processing: while the 4-phase feature selection pipeline (MI, RFE, importance ranking, and correlation filtering) ensures only relevant data is processed, the low latency SVM prediction makes it highly suitable for integration into browser extensions or edge level firewall APIs. |
ICACT20260103 Question.2
Questioner: hripsime.hovhannisyan@rau.am 2026-02-10 오후 4:44:20 |
ICACT20260103 Answer.2
Answer by Auhor arjunvirvashisht@gmail.com 2026-02-10 오후 4:44:20
Chrome Click!! |
| What is the inference latency of the model, and is it low enough to support real-time line-rate cyber defense? |
This is a critical point for practical cybersecurity. In our experimental setup, the average inference latency per URL is approximately 0.15 to 0.25 milliseconds on standard hardware. This efficiency is achieved because the final classification is offloaded to a Support Vector Machine (SVM) using the deep temporal embeddings extracted by the BiLSTM. For real-time line-rate defense, our framework supports asynchronous processing: while the 4-phase feature selection pipeline (MI, RFE, importance ranking, and correlation filtering) ensures only relevant data is processed, the low latency SVM prediction makes it highly suitable for integration into browser extensions or edge level firewall APIs. |
| |
