의사결정나무를 이용한 북한 탄도미사일 탄종 신속 분류 모델 = Rapid Classification Model of North Korean Ballistic Missile Types Using Decision Tree

국문 초록 (Abstract)

성공적인 대탄도미사일 작전을 수행하기 위해서는 북한의 탄도미사일 발사를 조기에 탐지하고, 해당 탄도미사일이 무엇인지 정확하게 초도판단하여 지휘관과 작전요원에게 신속하게 전파하는 것이 매우 중요하다. 본 연구는 북한이 발사한 탄도미사일의 탄종을 신속하고 정확하게 분류할 수 있는 머신러닝 모델을 제시한다. 데이터셋은 언론에 공개된 과거 북한 탄도미사일 발사정보와 시뮬레이션, 데이터 증강기법을 통해 확보하였다. 탄도미사일의 비행궤적을 결정하는 비행거리, 정점고도, 최고속도 등 최소한의 비행정보와, 북한의 도발 경향성을 고려하여 탄도미사일 발사장소를 변수로 활용한 의사결정나무 모델을 구축하였다. 층화 교차검증 등을 통해 모델의 일반화 성능을 평가한 결과, 균형 정확도(Balanced Accuracy) 95% 이상 수준의 우수한 성능을 달성하였다.

다국어 초록 (Multilingual Abstract)

To ensure the success of counter-ballistic missile operation, it is crucial to detect North Korean ballistic missile launches early and accurately identify the missile type to quickly relay this information to commanders and operational personnel. This study presents a machine learning model capable of rapidly and accurately classifying the types of ballistic missiles launched by North Korea. The dataset was con- structed using publicly available information from past North Korean ballistic missile launches, along with simulation and data augmentation techniques. The model was built using a decision tree algorithm, in- corporating minimal flight information such as flight distance, apogee, and maximum velocity, as well as the launch location, which reflects North Korea's provocative patterns. The model's generalization perform- ance was evaluated through stratified cross-validation, demonstrating an outstanding balanced accuracy of over 95%.

참고문헌 (Reference)

1 CNS, "The North Korean Missile Test Tracker" h t t p s : / / w w w . n t i . o r g / a n a l ysis/articles/cns-north-korea-missile-test-database/ 2024

2 장영근, "Technical Evaluation and Implications of North Korea's New Tactical Missile(Improved KN-23)" 28 (28): 123-138, 2021

3 Park, Y. S., "Statistical Machine Learning Using Python" Jayu Academy 2020

4 Sunny Kusawa, "SMOTE -Handle imbalanced dataset"

5 Tang, W., "Radar Target Recognition of Ballistic Missile in Complex Scene" 91-98, 2019

6 홍지수 ; 전세진, "Prediction of Safety Grade of Bridges Using the Classification Models of Decision Tree and Random Forest" 43 (43): 397-411, 2023

7 오주호 ; 강동수, "Predicting the Ballistic Missile Range using Long Short Term Memory" 28 (28): 405-412, 2022

8 Seungchan, S., "Missile from Alpha to Omega: The Core of K-Defense" Goodland 2023

9 He, H., "Learning from imbalanced data" 21 (21): 1263-1284, 2009

10 Tan, P. N., "Introduction to Data Mining" Addison-Wesley Longman Publishing Co., Inc 2005

11 한영진 ; 조인휘, "Imbalanced Data Improvement Techniques Based on SMOTE and Light GBM" 11 (11): 445-452, 2022

12 Gaiduchenko, N. E., "Hypersonic Vehicle Trajectory Classification Using Convolutional Neural Network" 123-128, 2019

13 Zhang, C., "Ensemble Machine Learning: Methods and Applications" Springer 2012

14 Singh, U., "Dynamic classification of ballistic missiles using neural networks and hidden Markov models" 19 : 280-289, 2014

15 Provost, F., "Data Science for Business: What You Need to Know about Data Mining and Data-Analytic Thinking" O'Reilly Media 2013

16 Lee, D. K., "Classification of Ballistic Missile Trajectories Using Univariate Time Series Classifier" 190-191, 2023

17 홍동욱 ; 임동순 ; 최봉완, "Application and Determination of Defended Footprint Using a Simulation Model for Ballistic Missile Trajectory" 21 (21): 551-561, 2018

18 Kim, J. W., "Analysis of the Flight Trajectory Characteristics of Medium-Range Ballistic Missiles by Adjusting Flight Path Angles" 23 (23): 35-44, 2015

19 유상은 ; 장대성, "Analysis of Defense Effectiveness in Cooperative Engagement for Multi-layered Missile Defense System With Differing Ballistic Missile Trajectories, Journal of Institute of Control" 29 (29): 671-678, 2023

20 유병천 ; 김주현 ; 권용수 ; 최봉완, "A Study on the Flight Trajectory Prediction Method of Ballistic Missiles-BM type by Adjusting the Angle of a Flight Path and a Range" 16 (16): 131-140, 2020

21 조성진, "A Study on a Simple Analysis Method for Rapidly Estimating the Trajectory of a Ballistic Missile" 49 (49): 1-11, 2023

22 김흥섭 ; 김기태 ; 전건욱, "A Requirement Assessment Algorithm for Anti-Ballistic Missile Considering Ballistic Missile's Flight Characteristics" 14 (14): 1009-1017, 2011

23 Ministry of National Defense Republic of Korea, "2022 Defense White Paper"