Polii, Sidney Cempaka (2024) Multifaceted analysis of missile attacks in Ukraine with machine learning: trends, impacts and future risks = Analisis multifaset serangan rudal di Ukraina dengan machine learning: tren, dampak, dan risiko di masa depan. Bachelor thesis, Universitas Pelita Harapan.
![Title [thumbnail of Title]](http://repository.uph.edu/style/images/fileicons/text.png)
Title.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (28kB)
![Abstract [thumbnail of Abstract]](http://repository.uph.edu/style/images/fileicons/text.png)
Abstract.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (192kB)
![TOC [thumbnail of TOC]](http://repository.uph.edu/style/images/fileicons/text.png)
TOC.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (806kB)
![Chapter 1 [thumbnail of Chapter 1]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter 1.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (858kB)
![Chapter 2 [thumbnail of Chapter 2]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter 2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (1MB)
![Chapter 3 [thumbnail of Chapter 3]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter 3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (968kB)
![Chapter 4 [thumbnail of Chapter 4]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter 4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (2MB)
![Chapter 5 [thumbnail of Chapter 5]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter 5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (370kB)
![Bibliography [thumbnail of Bibliography]](http://repository.uph.edu/style/images/fileicons/text.png)
Bibliography.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (436kB)
![Appendices [thumbnail of Appendices]](http://repository.uph.edu/style/images/fileicons/text.png)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (1MB)
Abstract
In unfamiliar situations, a deep understanding is crucial for determining the
appropriate response, including in the geopolitical conflict between Russia and Ukraine.
This study aims to explore the conflict further and analyze attack patterns using machine
learning. The dataset includes Russian attacks on Ukraine from October 2022 to July
2024.
This research encompasses several types of analysis, including predictive impact
analysis, geospatial analysis, and association analysis. The predictive impact analysis
employs logistic regression and random forest models. The results indicate that random
forest better captures complex data patterns, with an AUC of 0.56 compared to 0.5 for
logistic regression. However, logistic regression has a higher recall of 1.0 compared to
0.98 for random forest. Feature importance analysis from the random forest model
identifies three key factors influencing attacks: missile model category (Model Num),
missile carrier category (Carrier Num), and the number of missiles launched (Launched).
Geospatial analysis is used to map attack patterns through visualization with the
Pydeck library. Data processing involves merging location datasets with latitude and
longitude coordinates, requiring approximately three hours of processing time. The
results reveal several locations that are more frequently targeted in attacks. Meanwhile,
association analysis using the Apriori algorithm successfully identifies five strong
condition pairs from 1,421 analyzed data rows. A lift value greater than one indicates a
significant correlation between specific attack variables.
The findings of this study are expected to contribute to the development of more
effective defense strategies to minimize casualties and damage caused by attacks. / Dalam menghadapi situasi yang belum dikenal, pemahaman yang mendalam
sangat penting untuk menentukan respons yang tepat, termasuk dalam konflik geopolitik
antara Rusia dan Ukraina. Penelitian ini bertujuan untuk menggali lebih dalam peristiwa
tersebut serta menganalisis pola serangan menggunakan machine learning. Data yang
digunakan mencakup serangan Rusia terhadap Ukraina dari Oktober 2022 hingga Juli
2024.
Penelitian ini mencakup beberapa analisis mulai dari analisis prediktif dan
dampak, geospasial, dan asosiasi. Analisis prediktif dan dampak menggunakan model
regresi logistik dan random forest. Hasilnya menunjukkan bahwa random forest lebih
mampu menangkap pola data yang kompleks dengan nilai AUC 0.56 dibandingkan
regresi logistik yang hanya mencapai 0.5. Namun, regresi logistik memiliki recall lebih
tinggi sebesar 1.0 dibandingkan dengan 0.98 pada random forest. Analisis feature
importance dari random forest mengidentifikasi tiga fitur utama yang paling berpengaruh
dalam serangan, yaitu kategori model rudal (Model Num), kategori pembawa rudal
(Carrier Num), dan jumlah rudal yang diluncurkan (Launched).
Analisis geospasial digunakan untuk memetakan pola serangan dengan visualisasi
menggunakan pustaka Pydeck. Pengolahan data melibatkan penggabungan dataset lokasi
dengan koordinat garis lintang dan bujur, yang memerlukan waktu pemrosesan sekitar
tiga jam. Hasil analisis menunjukkan beberapa lokasi yang lebih sering menjadi target
serangan. Sementara itu, analisis asosiasi dengan algoritma apriori berhasil menemukan
lima pasangan kondisi dengan hubungan yang kuat dari 1.421 baris data yang dianalisis.
Nilai lift lebih dari satu menunjukkan bahwa pola serangan memiliki keterkaitan
signifikan antara variabel tertentu.
Diharapkan hasil penelitian ini dapat berkontribusi dalam pengembangan strategi
pertahanan yang lebih efektif guna mengurangi korban jiwa dan kerusakan akibat
serangan.
Item Type: | Thesis (Bachelor) |
---|---|
Creators: | Creators NIM Email ORCID Polii, Sidney Cempaka NIM0108221022 sdnycpolii@gmail.com UNSPECIFIED |
Contributors: | Contribution Contributors NIDN/NIDK Email Thesis advisor Samosir, Feliks Victor Parningotan NIDN1319049302 feliks.parningotan@uph.edu |
Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
Divisions: | University Subject > Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics |
Depositing User: | Stefanus Tanjung |
Date Deposited: | 09 Aug 2025 05:33 |
Last Modified: | 09 Aug 2025 05:33 |
URI: | http://repository.uph.edu/id/eprint/70417 |