Research on a Real-time Computer Network Trend Analysis Algorithm Based on Dynamic Data Flow under the Background of Big Data

With the continuous development of information technology, the Internet technology and computing technology are constantly improving and developing, and the demand of users is constantly increasing in this process, so the network specification is particularly important in this context. Traditional behavior identity is mostly based on a static model, which has a strong dependence on sample eigenvalues, but lacks flexibility and adaptability to individual users' behavior habits. In this paper, the data stream is segmented and fitted based on the total least square method, and the traditional curve analysis algorithm and online data segmentation are combined and improved. The variable sliding window algorithm is used to realize the reasonable segmentation of the data stream and improve the accuracy of trend analysis. Experimental results show that the algorithm has high recognition rate and good adaptability to users.