Risk prediction method of multi-source network information cross-space fusion based on reinforcement learning

In order to improve the level of cross-space transmission and sharing of multi-source network information, it is necessary to effectively predict the process risk of fusion, and a risk prediction method of cross-space fusion of multi-source network information based on reinforcement learning is proposed. A cross-spatial dimension construction and transformation model of multi-source network information is constructed, and a feature-tag correlation selection model is established. Combined with the maximum feature-tag correlation identification method, the cross-spatial transformation and reinforcement tracking learning of multi-source network information are realized, and the redundancy test in the process of information cross-spatial fusion is realized by reinforcement learning. The frame structure model of network topology transmission is established according to the self-organized decentralized distributed networking mode of multi-source network space, and the reinforcement learning of information fusion risk prediction is realized by using the replacement parameter configuration and protocol configuration of channel resources, according to the reinforcement learning results. The simulation results show that the proposed method has a high degree of fitness for risk prediction of cross-space fusion of multi-source network information, improves the utilization rate of networking channels, and has good reliability in the process of channel access and resource allocation, effectively reducing the fusion risk, thus improving the transmission channel balance of multi-source networks.