<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In this paper, there is no unified grading standard for the harm of terrorist attacks. A classification model of terrorist incidents based on machine learning is proposed. First, the data related to the hazard in the Global Terrorism Database (GTD) is extracted and preprocessed. Secondly, the data is extracted by principal component analysis, and all events are aggregated into 5 by K-means clustering. Again, the entropy method is used to calculate the weighting coefficient of each indicator, and the comprehensive score of the hazard of each type of terrorist attack is calculated. Finally, the scores are divided into 1-5 levels of hazard grading models in order of high to low. The results show that the hazard grading model can scientifically and objectively quantify terrorist attacks.</p>
</abstract>

In this paper, there is no unified grading standard for the harm of terrorist attacks. A classification model of terrorist incidents based on machine learning is proposed. First, the data related to the hazard in the Global Terrorism Database (GTD) is extracted and preprocessed. Secondly, the data is extracted by principal component analysis, and all events are aggregated into 5 by K-means clustering. Again, the entropy method is used to calculate the weighting coefficient of each indicator, and the comprehensive score of the hazard of each type of terrorist attack is calculated. Finally, the scores are divided into 1-5 levels of hazard grading models in order of high to low. The results show that the hazard grading model can scientifically and objectively quantify terrorist attacks.

Hazard Grading Model of Terrorist Attack Based on Machine Learning

International Journal of Advanced Network, Monitoring and Controls

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

{"article-title":"Hazard Grading Model of Terrorist Attack Based on Machine Learning"}

In this paper, there is no unified grading standard for the harm of terrorist attacks. A classification model of terrorist incidents based on machine learning...

Hazard level	Cluster category	Hazard level
1	2	1766.7104
2	3	3.2596
3	0	0.6239
4	4	-2.6904
5	1	-0.8788

Field	Description
extended	Whether it is a continuous event
latitude	latitude
longitude	longitude
success	Successful attack
suicide	Suicide attack
nkill	Total number of deaths
propextent	Degree of property damage
nwound	Total number of injuries
country	country
region	area
city	city
attacktype	Attack type
targtype	Target/victim type
weapontype	Weapon type

Indicators	Characteristic values
nkill	9.82022087e-01
nwound	8.06184462e-02
targtype	7.91122120e-03
country	5.20872985e-02
attacktype	4.84991077e-03
region	4.01240379e-02
suicide	2.66626688e+00
city	2.60031933e-02
longitude	1.84972981e+02
extended	1.63936354e+03
latitude	1.36725606e+03
propextent	1.06560032e-01
success	1.04574700e+02
weapontype	0.00000001e+00

Index	assignment
developed countries	2
underdeveloped countries	1
the capital	3
the provincial capital	2
other cities	1

type	X1	X2	X3	X4	numbers
0	2.4843	-16.3826	-1.3464	0.3081	63122
1	-3.3968	22.8297	-3.8782	0.0615	37848
2	825.778	873.697	28.9316	-104.59	2
3	13.8411	-127.794	19.7789	-2.7281	3500
4	-9.5985	63.3898	16.7324	-1.2382	9711

Hazard Grading Model of Terrorist Attack Based on Machine Learning

Data publikacji: 08 paź 2019

Zakres stron: 81 - 85

DOI: https://doi.org/10.21307/ijanmc-2019-051

Słowa kluczowe
Terrorist Attacks, Hazard, Hierarchical Model, Principal Component Analysis, K-Means Clustering, Entropy Method

© 2019 Yu Jun et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

WEIGHT COEFFICIENTS OF EACH INDICATOR

HAZARD GRADING RESULT

THE SELECTED FIELD TABLE

CHARACTERISTIC VALUES CORRESPONDING TO THE INDICATORS

THE STATE AND CITY ASSIGNMENT

CLUSTERING CENTER FOR EVENT CLASSIFICATION

Hazard Grading Model of Terrorist Attack Based on Machine Learning

Data publikacji: 08 paź 2019

Zakres stron: 81 - 85

DOI: https://doi.org/10.21307/ijanmc-2019-051

Słowa kluczoweTerrorist Attacks, Hazard, Hierarchical Model, Principal Component Analysis, K-Means Clustering, Entropy Method

© 2019 Yu Jun et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

WEIGHT COEFFICIENTS OF EACH INDICATOR

HAZARD GRADING RESULT

THE SELECTED FIELD TABLE

CHARACTERISTIC VALUES CORRESPONDING TO THE INDICATORS

THE STATE AND CITY ASSIGNMENT

CLUSTERING CENTER FOR EVENT CLASSIFICATION

Słowa kluczowe
Terrorist Attacks, Hazard, Hierarchical Model, Principal Component Analysis, K-Means Clustering, Entropy Method