Research on the Object Recognition of Minimum Living Standard Security System Based on Neural Network
PROCEEDINGS OF 2016 12TH INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE AND SECURITY (CIS)(2016)
摘要
This paper focuses on the problem of subjectivity and low-efficiency on the object identification of the Minimum Living Standard Security System (hereinafter referred to as MLSSS). Today, it is difficult to analyze the characteristics of the families which can or cannot pass the application of receiving allowances from the MLSSSs. To recognize whether those families are qualified automatically, we build a neural network model. The samples gathered from applicants for the MLSSS is in an extremely imbalance, because only few of them is unqualified. In order to improve the recognition accuracy of the Poor Families applying for the MLSSS, this article use oversampling technology to optimize the model. Applying the data mining techniques to the object recognition of MLSSS will both improve the accuracy and efficiency, and meanwhile, expand the applications of neural network technology.
更多查看译文
关键词
object recognition of Minimum Living Standard Security System, neural network, oversampling
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络