【摘要】 [目的/意义]高校网络舆情预测在高校应急管理中占据着重要的地位，正确的预测分析能够为舆情的引导和管理提供基础，对增强高校网络舆情应对能力、维护高校和谐稳定有着重要的作用和意义。[方法/过程]应用细胞型P系统的结构和操作，以及膜算法的规则，结合遗传算法设计动态算子，优化LSTM神经网络，得到基于动态膜驱动的长短期记忆神经网络预测模型(DM-LSTM),开展高校网络舆情预测研究。[结果/结论]提出一种基于动态膜驱动的长短期记忆神经网络预测模型(DM-LSTM),通过实证研究和对比分析，证明所提出的DM-LSTM模型在预测精度和算法性能方面具有更好的优越性和稳定性。更多还原

【Abstract】 [Purpose/significance] University network public opinion prediction plays an important role in university emergency management. Correct predictive analysis can provide a basis for the guidance and management of public opinion, and plays an important role and significance in enhancing the ability of universities to respond to network public opinion and maintaining the harmony and stability of universities.[Method/process]Applying the structure and operation of the cellular P system, as well as the rules of the membrane algorithm, combined with the genetic algorithm to design dynamic operators, the LSTM neural network is optimized, and the long and short-term memory neural network prediction model based on dynamic membrane drive(DM-LSTM) is obtained to carry out the research on the prediction of network public opinion in universities.[Result/conclusion]This paper proposes a long and short-term memory neural network prediction model based on dynamic membrane driving(DM-LSTM). Through empirical research and comparative analysis, it is proved that the proposed DM-LSTM model has better superiority and stability in terms of prediction accuracy and algorithm performance.更多还原