Final Exam Task
Military Artificial Intellegent
Lectures :
Prof. Dr. Eng. Wisnu Jatmiko, S.T., M.Kom.
Dr. Ario Yudo Husodo, S.T., M.T.
Grafika Jati, S.Kom., M.Kom.
Group 6 :
Andri Purwoko – 120220405005
Gunawan Wibisono - 120220405010
Hetly Brinner Yussak Kawet - 120220405012
John Swatrahadi Permana -120220405013
PRODI REKAYASA PERTAHANAN SIBER
FAKULTAS TEKNOLOGI PERTAHANAN
UNIVERSITAS PERTAHANAN REPUBLIK INDONESIA
2022
General Task
Explore the Usage of Artificial Intelligence Code related to the Military field.
Analyze the code used to build the AI.
Modify the code to improve its performance.
The Final Exam Task is group-based.
Specific Task
a) Split the participants of the Military Artificial Intelligence Course into six groups
with a relatively equal number of members.
b) Explore the code related to artificial intelligence provided in:
1. https://paperswithcode.com/task/network-intrusion-detection/codeless
2. https://github.com/rahulvigneswaran/Intrusion-Detection-Systems
c) Each group in point (a) should choose one of the codes, whether b1 or b2, to
be analyzed.
d) Each group should produce one analytics document consisting of 15-25 pages
(excluding the cover and table of content) that discuss the code chosen in step (c).
e) In the document of point (d), each group should:
1. Describe the problem discussed in b1 or b2.
2. Describe why Artificial intelligence is needed to solve e1.
3. Choose one of the classical approaches alternative available to solve e1.
Describe how it works. For point e3, no two different groups can choose the
same approach.
4. Choose one of the deep learning approaches alternative available to solve e1.
Describe how it works. For point e4, it is okay for two different groups to choose
the same approach.
5. Improve the algorithm's performance in steps e3 and e4 by modifying the
available source code. Each group should describe why and how they change
the code, then provide the improvement result as a chart in the document of
point (d).
f) Each group can construct the document structure in point (d) as they wish.
Nevertheless, each group should consider that the crucial score of this task is ordered
like this: element e5 > e4 > e3 >e2 > e1.
Introduction databaset
We choose : UNSW-NB15 dataset with the paper first winner that using Edge-
Detect-FRNN in 2021 with accuracy 99,6 %
The raw network packets of the UNSW-NB 15 dataset was created by the IXIA
PerfectStorm tool in the Cyber Range Lab of UNSW Canberra for generating a hybrid of
real modern normal activities and synthetic contemporary attack behaviours. The
tcpdump tool was utilised to capture 100 GB of the raw traffic (e.g., Pcap files). This
dataset has nine types of attacks, namely :
1. Fuzzers
2. Analysis
3. Backdoors
4. DoS
5. Exploits
6. Generic
7. Reconnaissance
8. Shellcode and
9. Worms.
The Argus, Bro-IDS tools are used and twelve algorithms are developed to
generate totally 49 features with the class label. These features are described
in UNSW-NB15_features.csv file.
The total number of records is two million and 540,044 which are stored in the
four CSV files, namely, UNSW-NB15_1.csv, UNSW-NB15_2.csv, UNSW-NB15_3.csv
and UNSW-NB15_4.csv. The ground truth table is named UNSW-NB15_GT.csv and the
list of event file is called UNSW-NB15_LIST_EVENTS.csv. A partition from this dataset
was configured as a training set and testing set, namely, UNSW_NB15_training-set.csv
and UNSW_NB15_testing-set.csv respectively. The number of records in the training
set is 175,341 records and the testing set is 82,332 records from the different types,
attack and normal.
Free use of the UNSW-NB15 dataset for academic research purposes is hereby
granted in perpetuity. Use for commercial purposes should be agreed by the authors.
Nour Moustafa and Jill Slay have asserted their rights under the Copyright. To whom
intend the use of the UNSW-NB15 dataset have to cite the above five papers.
1. The problem discussed
- Cyber attack growing
- light => computing power
- fast => cost drop
- and accurate
2. Artificial intelligence is needed to solve
-
3. one of the classical approaches alternative available to solve, how its work
- Traditional methods of intrusion detection and deep packet inspection, while
still largely used and recommended, are no longer sufficient to meet the
demands of growing security threats
4. one of the deep learning approaches alternative available to solve
5. Improve the algorithm's performance, why and how they change the code
6. then provide the improvement result as a chart
