Network traffic prediction method and system for improving lightning connection process optimization algorithm

Abstract

The invention belongs to the technical field of network traffic prediction, and discloses a network traffic prediction method and system for improving a lightning connection process optimization algorithm, and the method comprises the steps: importing network traffic data into a model through a pandas tool, carrying out the data cleaning and data enhancement, dividing the data into a training set, a verification set and a test set, determining the number of units of an input layer, an output layer and a hidden layer in the neural network; optimizing a search policy Levy-LAPO of an LAPO algorithm in combination with a Levy flight and lightning connection process so as to search an optimal hyper-parameter of the neural network; and according to the searched optimal parameter, taking the optimal parameter as an initial parameter of the neural network for training, and obtaining the trained neural network. The method is low in cost and easy to popularize; the calculation amount is reduced, the convergence speed is high, and consumption and time consumption of calculation resources are saved; manual domain knowledge intervention is not needed, the application range of the domain is expanded, and the convergence performance of the algorithm is improved.

Description

technical field [0001] The invention belongs to the technical field of network traffic forecasting, and in particular relates to a network traffic forecasting method and system for improving an optimization algorithm of a lightning connection process. Background technique [0002] At present, with the increasing size and complexity of the network, the network bandwidth has increased sharply, and the network-based Internet services have shown diversified development, which has driven many new industries and network products, which makes the operation mechanism and behavior characteristics of the Internet complicated. It has brought a lot of jobs, but it has also brought about a huge demand for network traffic usage, and at the same time, the bandwidth demand for network traffic has also been unevenly distributed. Therefore, it has become a problem to be solved to quickly solve the limit of the system network traffic load capacity and to control and manage the network QoS in r...

AI Technical Summary

Problems solved by technology

However, the cyclic neural network takes too long, and the network communication only reads and parses the input load data once, so the deep neural network must wait for the previous input to be processed before the next input can be processed. Redundancy, and requires a lot of memory during training

[0013] (3) When using the deep learning model as the fitness function, there is a problem that the search speed of the optimal parameters is slow. At the same time, the computational complexity is high, the optimization process has certain redundancy, and the convergence speed is slow.

[0017] (3) Large-scale and high-latitude data collection is difficult, and data standardization is time-consuming and labor-intensive

[0018] (4) It is difficult to overcome the shortcomings of optimization algorithms that are easy to fall into local optimum and premature fitting

Images