Micro Expression Recognition by Machine Learning Based Profit Function Analysis in Intelligent Marketing of Financial Industry

Microexpression recognition (MER) has wide applications in the financial industry. MER is based on machine learning or deep learning algorithm to establish a classification model to capture the subtle expression changes of the face from the video. We propose a hybrid intelligent marketing scheme (HIMS). In the HIMS, firstly, the system automatically recommends products to customers according to the Big data analysis results, while monitoring the emotional changes of customers in real-time. When the system detects that customers have negative emotions, the marketing scheme turns to manual mode. The key step of HIMS is to monitor the emotional changes of customers through the MER model. The advantage of HIMS is that it can reduce marketing costs while ensuring a larger marketing scale, higher marketing efficiency, and lower complaint rate. Based on the real marketing scenario, we deduce the profit function and give the analysis results. The experimental results show that under multiple parameter configurations, the profit of HIMS is greater than that of the Big data marketing scheme with an increase of 8.8-12.8 times. For products with more customer complaints, this advantage becomes significant and is insensitive to commodity value and has strong robustness.