運用決策樹預測與預防台灣高犯罪風險青少年之犯罪問題

Using Decision Tree to Predict and Protect Crime Problem of High Offense Risk Adolescent in Taiwan

陳詠霖(Yung-Lin Chen)；郭玟君(Min-Jyun Guo)

青少年犯罪 ； 青少年教育 ； 行政管理 ； 決策樹 ； Adolescent Crime ； Adolescent Education ； Administration Management ； Decision Tree

聯大學報

12卷2期（2015 / 12 / 01）

139 - 149

繁體中文

台灣當前正面臨少子化問題，所以需要妥善照顧、管理與教導青少年以利台灣未來可以獲得優質人力與豐厚稅收；為了在政府有限資源下優先輔導比較可能產生問題之青少年以達到「事半功倍」之成效，本研究使用C5.0 機器學習技術進行青少年犯罪可能性之分類預測，並運用此分類技術產生分類規則以提供政府進行青少年輔導與管理之建議；本研究先探討台灣青少年犯罪和機器學習相關文獻並介紹C5.0 決策樹分類機，然後收集台灣青少年犯罪的相關資料並進行分類，並將C5.0 決策樹分類機之分類結果與C4.5 分類機、迴歸分析分類機、K 鄰近分類機、類神經網路分類機、PNN 分類機之分類結果進行比較，根據實驗結果，C5.0 決策樹分類機之分類成效優於上述其他分類機置，根據C5.0 決策樹所產生之分類規則，本研究建議政府應關懷失學青少年、嚴格規範青少年出入場所、提供弱勢青少年經濟支援以降低青少年犯罪之可能性。

Taiwan faces the problem of fewer children now. So, Taiwan government should take care, manage and educate adolescent to acquire high quality manpower and plentiful taxes in the future for Taiwan. In this limited government resource environment, government should educate the high offense risk adolescent first to maximize the education and management performance in crime problem of adolescent. This research use C5.0 machine learning technique to classify and predict the possibility of adolescent crime to generate some classification rule for providing government some suggestion about adolescent tutorship and management. This research collects some literature about adolescent crime problem and machine learning. This study also introduces C5.0 decision tree classifier. And then, the relative information of Taiwan adolescent will be collected and should be generated some classification rule to classify by C5.0 classifier. The classification performance of C5.0 classifier will be compared with C4.5 classifier, regression classifier, KNN classifier, BPNN classifier and PNN classifier. According to experiment result, the classification performance of C5.0 classifier is better than all of the other classifiers. According to the classification rule which is generated by C5.0 classifier, this research suggests that government should take care of adolescents who do not study in school, limit the place where the adolescents cannot enter seriously and support the adolescent whose economics ability is worse. All of above mechanism can reduce the possibility of adolescent crime.

1. Cheng, D.,Wang, J.,Wei, X.,Yihong Gong, Y.(2015).Training mixture of weighted SVM for object detection using EM algorithm.Neurocomputing,149,473-482.
2. Garca-Laencina, P. J.,Sancho-Gomez, J. L.,Figueiras-Vidal, A. R.,Verleysen, M.(2009).K nearest neighbours with mutual information for simultaneous classification and missing data imputation.Neurocomputing,72,1483-1493.
3. Giri, D.,Acharya, U. R.,Martis, R. J.,Sree, S. V.,Lim, T. C.,VI, T. A.,Suri, J. S.(2013).Automated diagnosis of Coronary Artery Disease affected patients using LDA, PCA, ICA and Discrete Wavelet Transform.Knowledge-Based Systems,37,274-282.
4. Kobayashi, D.,Takahashi, O.,Arioka, H.,Koga, S.,Fukui, T.(2013).A prediction rule for the development of delirium among patients in medical wards: Chi-Square Automatic Interaction Detector (CHAID) decision tree analysis model.The American Journal of Geriatric Psychiatry,21(10),957-962.
5. Li, D. F.,Hu, W. C.,Xiong, W.,Yang, J. B.(2008).Fuzzy relevance vector machine for learning from unbalanced data and noise.Pattern Recognition Letters,29,1175-1181.
6. Mousavi, R.,Eftekhari, M.(2015).A new ensemble learning methodology based on hybridization of classifier ensemble selection approaches.Applied Soft Computing,37,652-666.
7. Ostermark, R.(2009).A fuzzy vector valued KNN-algorithm for automatic outlier detection.Applied Soft Computing,9,1263-1272.
8. Pai, P. F.,Li, L. L.,Hung, W Z(2014).Using ADABOOST and Rough Set Theory for Predicting Debris Flow Disaster.Water resources management,28(4),1143-1155.
9. Pang, S. L.,Gong, J. Z.(2009).C5.0 classification algorithm and application on individual credit evaluation of banks.Systems Engineering-Theory & Practice,29(12),94-104.
10. Polat, K.,Güneş, S.(2009).A novel hybrid intelligent method based on C4. 5 decision tree classifier and one-against-all approach for multi-class classification problems.Expert Systems with Applications,36(2),1587-1592.
11. Quinlan, J. R.(1993).C4.5: Programs for Machine Learning.Morgan Kaufmann Publishers.
12. Tian, D.,Zeng, X. J.,Keane, J.(2011).Core-generating approximate minimum entropy discretization for rough set feature selection in pattern classification.International Journal of Approximate Reasoning,52(6),863-880.
13. Tsai, C. F.,Hsu, Y. F.,Yen, David C.(2014).A comparative study of classifier ensembles for bankruptcy prediction.Applied Soft Computing,24,977-984.
14. Tsai, C. Y.,Chen, C. J.(2015).A PSO-AB classifier for solving sequence classification problems.Applied Soft Computing,27,11-27.
15. Yasami, Y.,Mozaffari, S. P.(2010).A novel unsupervised classification approach for network anomaly detection by k-Means clustering and ID3 decision tree learning methods.The Journal of Supercomputing,53(1),231-245.
16. Yu, L.,Yao, X.,Wang, S.,Lai, K. K.(2011).Credit risk evaluation using a weighted least squares SVM classifier with design of experiment for parameter selection.Expert Systems with Applications,38(12),15392-15399.
17. Zhang, Y.,Meratnia, N.,Havinga, P. J. M.(2013).Distributed online outlier detection in wireless sensor networks using ellipsoidal support vector machine.Ad Hoc Networks,11,1062-1074.
18. 江小燕(2004)。碩士論文(碩士論文)。國立政治大學心理學研究所。
19. 周國雄(2004)。碩士論文(碩士論文)。中央警察大學犯罪防治研究所。
20. 林秀怡(2010)。博士論文(博士論文)。中央警察大學犯罪防治研究所。
21. 林惠鈴(2006)。碩士論文(碩士論文)。中國文化大學青少年兒童福利研究所。
22. 施佩姍(2011)。碩士論文(碩士論文)。國立政治大學財政研究所。
23. 商嘉昌(2006)。碩士論文(碩士論文)。國立政治大學社會學系。

1. 黃詩淳、邵軒磊(2018)。酌定子女親權之重要因素：以決策樹方法分析相關裁判。臺大法學論叢，47(1)，299-344。