Time series data prediction method and device and computer readable storage medium

Abstract

The invention provides a time series data prediction method and device and a computer readable storage medium, and relates to the technical field of computers. The method comprises the steps of receiving time sequence data of a to-be-tested user; inputting the time sequence data into a target classification network of the trained time sequence mode so as to enable the target classification networkto perform dimension reduction compression on the time series data to obtain corresponding derivative feature data, performing target classification on the derivative feature data to obtain a prediction probability of a target classification result corresponding to the time series data, and outputting the prediction probability; and determining a test result of the to-be-tested user based on theprediction probability. According to the embodiment of the invention, the prediction probability of the obtained target classification result of the application time point is ensured to be consistentwith the distribution of the target classification result of the modeling time point as much as possible, so that a strategy constructed based on the trained time sequence model can run more stably and enduringly, and the actual application requirements are better met.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular, the present application relates to a time series data prediction method, device and computer-readable storage medium. Background technique [0002] With the popularization of mobile smart devices and the rapid development of the Internet, users can easily realize online activities such as communication, entertainment, payment, etc., as well as experience online and offline O2O life services simply by operating applications on mobile smart devices. Offline consumption such as navigation, private car, etc., basically covers most of the user's life trajectory, resulting in a wealth of user behavior data, so that based on these data, users can be more comprehensively described. [0003] However, due to the rapid development of the Internet, business adjustments and changes often occur, and even business offline directly. Because of this uncertainty, this leads to the i...

AI Technical Summary

Problems solved by technology

[0007] This application provides a time series data prediction method, device and computer-readable storage medium, which can solve the problem that the predicted value of the target classification result does not conform to the pre-built strategy

Images