Anomaly Detection by Time Series Motif Discovery for Vietnamese Customs  
Author Dong Van Hoang

 
Co-Author(s) Quang Anh Ngoc Pham; Nam Thanh Vu; Ngoc Anh Thi Nguyen

 
Abstract Financial transactions exploded recently by mobile, online transactions in the whole world. Fraud detection problems have been important great interests of many researchers. In data analytics, fraud detection is used machine learning methods. This paper proposes a supervised method by time series motif discovery for customs. A new proposed model is given using time series, motif discovery and machine learning for the real-life data of account company from customs data. The results of this paper achieve the new models for motif discovery and experiment results.

 
Keywords Time series, abnormal detection, motif discovery, machine learning, customs
   
    Article #:  DSBFI19-11
 
Proceedings of ISSAT International Conference on Data Science in Business, Finance and Industry
July 3-5, 2019 - Da Nang, Vietnam