Research on Trading Enterprise Risk Assessment using Channel Mapping based Butterfly Optimization Algorithm with Feed Forward Neural Network

Financial institutions have ability to invent time and geographical constrictions that speed up the business processing according to Internet technology. This provides quick financial services on an Internet for customers through financing needs. Since, the rise of finance and trading enterprises has also generated issues like user fraud and credit risk and also the existing risk assessment models failed to evaluate accurate results due to dynamic trade conditions. To overcome this limitations, a Channel Mapping based Butterfly Optimization Algorithm with Feed Forward Neural Network (CBOA-FFNN) is proposed for accurate prediction of risk levels and utilized for trading enterprise risk assessment. The main advantage of proposed FFNN model is that it reduces errors while predicting risk in trade enterprises by initializing the BOA model to explore risk levels accurately. The data acquired by different factors using block chain to improve the risk assessment process and preprocessed by removing outliers and filling missing data. Experimental results of the proposed CBOA-FFNN method in trading enterprise risk assessment attained achieved accuracy of 0.892 which is better results when compare to the existing methods such as regression technique and FFNN network.