网络信息传播最大化是当前互联网领域的重要研究方向之一。就网络信息传播最大化问题而言，现有的相应解决方法还不能很有效应对现实大规模、动态互联网络环境，它们的性能尚有待进一步改进提升。学习自动机LA作为一种强化机器学习技术，由于其具有全局优化、计算收敛快、自适应动态随机环境等优点，有望用来有效解决上述网络环境下的信息传播最大化问题，但是目前这方面的研究尚处于起步阶段，研究成果很少且性能上也存在一定不足。鉴于此，项目以研究适合解决大规模、动态互联网络环境下信息传播最大化问题的相应LA新算法为主要目标，具体研究多学习机制融合的快速学习自动机算法、基于跟踪技术和迁移学习的学习自动机算法、多学习自动机联合优化学习算法及基于学习自动机的网络信息传播最大化方法。预期成果可望在网络信息传播安全管理、网络社交、网络商业活动等领域具有良好的应用前景，也可望对机器学习等相关领域的研究在理论上起到促进作用。

Maximum information diffusion is one of the most important issues in the field of Internet technology. In the case of maximum network information diffusion problem, the existing approaches for maximum information diffusion cannot work very well when dealing with large-scale, dynamic Internets and still need to be improved to promote their performance. Learning automata (LA) is a kind of reinforcement learning technology. Because of its numerous advantages such as abilities of global optimization, fast convergence, self-adaptive to dynamic random environments, etc., LA is expected to provide an efficient way to solve maximum network information diffusion problem under such network environments above. However, the related research has just begun recently and there are only a little research outcomes with limited performance. Based on the above, the aim of the project is to study specific-designed LA algorithms that are fit for solving the maximum information diffusion problem under large-scale，dynamic Internet environments, and the major research contents include: fast LA algorithms based on multiple LA mechanism combination , LA algorithms based on tracking and transfer learning, multiple LA joint optimization algorithms, and LA based solutions for maximum network information diffusion. The expected outcomes of the project will have a wide application prospects in such fields as security management of network information diffusion, network socialising, network marketing, etc., and also will contribute to the theoretic development of the related research fields such as machine learning, etc.