Influenza prediction method based on image deep learning network integration model

The invention discloses an influenza prediction method based on an image deep learning network integration model, and belongs to the technical field of epidemic disease monitoring. Aiming at the problem that the prediction precision is influenced due to the fact that an existing method has limitation in the aspect of feature extraction and deep information in a sequence cannot be effectively mined, firstly, a one-dimensional digital sequence is coded into a two-dimensional image sequence based on a Grubrum angle field, image-based unique information features are provided for a model, and on the basis, the prediction precision of the model is improved; a deep learning network model is designed and built by integrating a sparrow search optimization algorithm, and feature extraction is performed by using the image processing capability of a convolutional neural network to complete flu prediction. A one-dimensional influenza sequence image code is introduced, and a nonlinear influenza prediction model optimized b