AI for Finance (AIFF): from Abnormal Data Recognition to Information System Intrusion Detection

With the development of the data mining and the intelligent information technology, the complex information processing paradigm based on artificial intelligence has become a mainstream trend. In this paper, starting from the discussions of abnormal data detection, the novel information system intrusion detection is designed for the construction of the AIFF model. The model starts with the recognition accuracy improvement, in the designed scheme, the weighted Naive Bayes algorithm and the intelligent clustering model are combined to effectively detection the abnormal information. Then, by adopting the rule that modification of the firewall rules on the proxy server makes it possible to directly filter out the next intrusion address from the firewall, the pre-processing will be achieved. Furthermore, this study proposes a scheduling strategy based on resource availability to finalize the detection system. Through the experiment under different scenarios of the self and also comparison testing, the performance is validated.