Computer intelligent prediction model of internet impulsive reward based on user portrait algorithm

In order to improve the matching degree between the prediction results of minors' impulsive reward behavior and the actual results, the design and research of minors' impulsive reward behavior prediction method based on user portrait is carried out. Take a webcast platform as an example to obtain the impulse reward data of some minors; Combined with the data, the attribute tag statistical database of minor users is constructed to describe the abnormal reward behavior of minors; Combined with the use of association rules, the trend of minors' impulsive reward is predicted, so as to complete the design of the prediction method of minors' impulsive reward behavior. The experimental results show that the designed method can accurately predict the impulsive reward behavior of minors in practical application, and the accuracy rate of prediction results is more than 0.90.