Improved Data Analysis Algorithm based on Multi-feature Network Construction

Abstract

Network information has become an important factor in today's social environment and network environment. With the large coverage of network data traffic and the new generation of network technology, illegal network data is also constantly invading the network environment, which has caused a serious network security threat. Therefore, the analysis and research of network data and the pre-judgment of feature types are of great significance. Based on the existing theoretical techniques, this paper proposes a Data Processing Algorithm Based on The Website Coding Structure (DPA) and a Multi-feature Network Construction Improvement Algorithm (MCIA), processing network data and analysis type prediction features. Experiments show that the data processing algorithm based on the website coding structure has the advantage of targeted processing of website data. It also verifies the validity of the proposed Digital Neighborhood Feature Set Algorithm (DNFA) and the improved algorithm for multi-feature network construction for data analysis. Application, compared with the existing single feature set, the pre-judgment basis for constructing multiple feature sets is more reliable, and the pre-judgment basis is more credible.

A Graph Memory Neural Network for Sea Surface Temperature Prediction

“…To fully consider spatial information, researchers generally adopt two approaches. The first one is to use spatial data, such as latitude, longitude, and regional features, as input for the model [21,22]. The second approach is to employ convolutional neural networks (CNN) to extract spatial features at different scales, and integrates them with time series prediction models to form a comprehensive spatiotemporal forecasting method [23][24][25].…”