Heat supply database query optimization method based on deep reinforcement learning

The invention belongs to the technical field of information retrieval, and discloses a heat supply database query optimization method based on deep reinforcement learning, which is characterized in that a view selection process is constructed as a Markov decision process, an intelligent agent and an environment are interacted by utilizing the deep reinforcement learning, learning is carried out in a'trial and error 'mode, and the search efficiency is improved. According to the method, the mapping strategy from the state to the action is continuously modified, the optimal solution of the materialized view is interactively and dynamically generated, the selection accuracy is improved, the selection speed of the materialized view is increased, and finally query optimization of the database is achieved. 本发明属于信息检索技术领域，公开了一种基于深度强化学习的供热数据库查询优化方法，本发明提出将视图选择过程构建为马尔可夫决策过程，利用深度强化学习让智能体与环境进行交互，以"试错"的方式进行学习，不断修改从状态到动作的映射策略，交互式地动态产生物化视图的最优解，既提高了选择的准确性，又提高了物化视图的选择速度，最终实现数据库的查询优化。