Research On Financial Risk Control Model And Algorithm Based On Machine Learning

With the development of Internet technology and the widespread use of applications,the use of the Internet for financial management has become a new type of financial technology.However,while Internet finance brings convenience,speed and high reporting rate to financial enterprises,it also brings considerable risks.Mainly manifested in:chaotic management,bad loans and malicious deception.Therefore,it is imperative to use the network operator big data to analyze the user behavior and avoid the above risks.Based on the above requirements,this paper studies the user behavior analysis algorithm based on operator big data to provide financial risk control mechanism for financial enterprises.Including:application data preprocessing,data portrait and label characterization,user behavior prediction,etc.Among them,the operator adopts the Unicom network platform as the basic platform and the financial APP as the data acquisition object.In the user behavior prediction algorithm,we adopt the Bayesian network as the prediction model,and build the risk prediction probability graph model by adding the "risk" label.The experimental analysis and tool prototype implementation of the proposed algorithm were carried out.Using the Unicom operator big data support platform,the financial APP online log as the acquisition object,and the SPARK parallel big data processing system as the analysis data environment,the experiment shows that the success rate of bad user behavior prediction has achieved satisfactory results,and the developed gadgets are in the enterprise.It has also been applied.