| abstract |
The invention provides a data grading storage algorithm based on an improved RBF neural network: the data stored for the first time is classified, the storage level is obtained according to the performance and capacity characteristics of the primary, secondary and tertiary storage devices, and the data is stored according to the grading result. When the data hierarchical storage system meets the migration conditions, the data migration factor value is calculated, a data hierarchical neural network model is established, and the mapping relationship between the data migration factor value and the storage level is obtained. The data migration factor value is used as the input of the data classification neural network model, the migration method is selected according to the trigger condition, and the migration data algorithm is screened according to the difference value P. The invention is oriented to the hierarchical storage strategy and model of multi-source, foreign, cross-system and multi-type data, realizes data migration between different levels of data, effectively improves data access efficiency and database utilization efficiency, improves data management decision-making efficiency, and accelerates Platform storage performance and reduce platform storage costs. |