Abstract | ||
---|---|---|
Research of the edge blur threshold recognition technology in big multimedia data classification has a great significance, which improves the data storage and safety performance. The traditional suspected boundary problem processing method mainly classified data through their features which were large amount, various types, less density of value and high speed of demand processing. That led to the problems such as inaccuracies and great errors. However, the edge blur threshold recognition technology summarized the methods of classifying data and put forward the principle of data classification. It classified the big multimedia data based on the reduction of feature dimensions and on the differences between the selected data. To determine the edge blur threshold, it used the least squares method. Combined with the decision tree method, it finally realized the classification of big multimedia data. The experimental results showed that the improved method has high precision and low recall rate with less time. This means the presented method has a certain advantage when compares with the classical method. |
Year | DOI | Venue |
---|---|---|
2018 | https://doi.org/10.1007/s11036-017-0926-6 | MONET |
Keywords | Field | DocType |
Big data,Fractal,Edge blur threshold,Data storage,Boundary problem,Decision tree | Least squares,Data mining,Decision tree,Recall rate,Computer science,Computer data storage,Fractal,Boundary problem,Data classification,Big data | Journal |
Volume | Issue | ISSN |
23 | 2 | 1383-469X |
Citations | PageRank | References |
0 | 0.34 | 24 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jia Wang | 1 | 79 | 17.75 |
Shuai Liu | 2 | 485 | 34.33 |
Houbing Song | 3 | 1771 | 172.26 |