Mobile video lag prediction model training method and prediction method

Abstract

The invention provides a mobile video lag prediction model training method. The method comprises the following steps: S1, obtaining a plurality of historical mobile video data with known lag conditions and time sequence to form a data set; s2, adopting the data set obtained in the step S1 to train the initial model to convergence according to the following mode: S21, dividing the data set into a training set and a test set; s22, dividing data in a training set obtained from the historical mobile video data into a plurality of training units according to autocorrelation; s23, training the model by using each training unit until the prediction accuracy of the model is greater than or equal to a preset accuracy threshold value or the training reaches a preset number of iterations so as to obtain a plurality of unit models, and when the LSTM network is trained by using the training units, performing correction weight updating by using a dropout function; and S24, taking the proportion of the prediction accuracy of each unit model on the test set in the sum of all prediction accuracy as a corresponding weight, and combining all the unit models to obtain a final mobile video lag prediction model.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular, to mobile video transmission jam prediction and corresponding resource allocation technology, and more particularly, to a mobile video jam prediction model training method and prediction method. Background technique [0002] With the rise of online video streaming services on the Internet and the influx of many short video content creators, the mobile short video market has begun to develop in a refined and vertical direction. Business has gradually become the mainstream business of network users. More and more video content is broadcast to the public on different platforms. More and more users are participating in the creation and browsing of mobile videos. Ensuring the smoothness of mobile videos is the basic ability of the network. The smooth playback of mobile video can greatly improve the user experience. Therefore, understanding and predicting the QoE (Qualit...

AI Technical Summary

Problems solved by technology

The prediction of mobile video freezing is a prediction problem related to time series. Therefore, the LSTM network has good applicability for the prediction of mobile video freezing. However, for the prediction of mobile video freezing, if the traditional LSTM network is directly used , will also cause problems such as overfitting and insufficient prediction accuracy, which will affect the subsequent resource allocation and other processes

Therefore, solving the problems of overfitting and insufficient accuracy of the traditional LSTM network is of great significance for the prediction of mobile video freezing and corresponding resource allocation

Images