ICACT20220392 Question.1
Questioner: tomayoon@ieee.org 2022-02-15 오전 10:12:50 |
ICACT20220392 Answer.1
Answer by Auhor shenhanji@cnic.cn 2022-02-15 오전 10:12:50
Chrome  Click!! |
| First of all, thank you for the good research paper and presentation contents. I would like to know more about the research environment used in this study, and also in detail about the equipment & operating systems, development languages and software libraries you used. |
Thank you for your question.The experiment environment adopts Huawei server Fusion Server G5500 (CentOS 7.6 system, Intel Golden 6230 Processor, 384Gb memory, Tesla V100 GPU Card), python development language and scikit-learn model library. |
ICACT20220392 Question.2
Questioner: tomayoon@ieee.org 2022-02-15 오전 10:21:14 |
ICACT20220392 Answer.2
Answer by Auhor shenhanji@cnic.cn 2022-02-15 오전 10:21:14
Chrome Click!! |
| Will you tell us what is the major achievement of this paper, and also what is further study issue from this? Thanks in advance!! |
This paper proposed a meta learning classification method based on few-shot samples, called BiLSTM-Ridge regression model, which can recognize and learn the characteristics of high threat log samples through less sample learning, and has better generalization detection ability in new unknown test tasks.
Future work includes applying this model to other fields, including high threat sample detection and classification of unmarked logs, and exploring the migration learning of other massive log detection in the cloud environment. |
ICACT20220392 Question.3
Questioner: ruigelwang@163.com 2022-02-16 오전 9:22:42 |
ICACT20220392 Answer.3
Answer by Auhor shenhanji@cnic.cn 2022-02-16 오전 9:22:42
Chrome Click!! |
| Have you used this BILSTM-Ridge Meta Learning Model to construct prototype system to testify this model? What was the effect? |
Thank you very much for your question. Since we still need to solve the task of Few-shot logs classification of other source data, we have not systematized the model. |
ICACT20220392 Question.4
Questioner: yanyonghang@henu.edu.cn 2022-02-16 오전 9:19:20 |
ICACT20220392 Answer.4
Answer by Auhor shenhanji@cnic.cn 2022-02-16 오전 9:19:20
Chrome Click!! |
| What is the space complexity and time complexity of the proposed scheme compared with LSTM Meta learning model, model independent MAML Meta learning model and relational network? |
Thank you very much for your question. These methods are the mainstream methods to solve the classification of Few-Shot samples. Due to the large difference of algorithm models and the small sample analysis, we pay more attention to the improvement of the accuracy of C way K shot task. We have not compared the spatial complexity and time complexity in the process of experiment. |
ICACT20220392 Question.5
Questioner: namacabale@gmail.com 2022-02-16 오후 3:40:00 |
ICACT20220392 Answer.5
Answer by Auhor shenhanji@cnic.cn 2022-02-16 오후 3:40:00
Chrome Click!! |
| Dear Author, you have a very good research. Is there an existing technology almost similar to your proposal that has been deployed in real world environment? What does it lack that your work specifically addressed? Thank you and stay safe. |
Good question！In real world environment, there are often only a small amount of labeled data in a certain field, but there are a large number of unlabeled data. Due to the differences of data sets, the methods used are different. There are many application scenarios in the CV field. At present, it is less used in high threat log analysis. |
