In the enterprise risk identification project, the data has the characteristics of the complex data sources and the diverse risks. The traditional feature extraction methods cannot cope with the multiple complex data sources at the same time. In this study, the risk feature extraction based on the single relationship of thein-depth learning and the feature coding technology of the risk atlas considering the time series can effectively extract the feature vectors of the risk atlas and avoid them at the same time. It avoids the complexity and blindness of the artificial feature extraction. And it is verified by a large-scale organization’s experiment on the enterprise’s overdue risk. Under the same model framework, the characteristics generated by using the risk feature coding technology are better than those given by the traditional experts under the model’s various indicators, which provide the strong support for the follow-up risk identification.
CITATION STYLE
Liu, D., & Qiu, Z. (2020). The Coding Technology of Operation and Management Risk Characteristics Based on the In-Depth Learning. In Advances in Intelligent Systems and Computing (Vol. 1088, pp. 1661–1671). Springer. https://doi.org/10.1007/978-981-15-1468-5_198
Mendeley helps you to discover research relevant for your work.