Research Article
Dynamic Prediction Research of Silicon Content in Hot Metal Driven by Big Data in Blast Furnace Smelting Process under Hadoop Cloud Platform
Table 3
Statistical table of dynamic prediction of the [Si]% value of three different algorithms.
| Error range | SVM algorithm for the serial Pegasos solver | The parallel SVM algorithm for Pegasos | SVM algorithm for the structural parallel Pegasos solver | Hit rate | The cumulative hit rate | Hit rate | The cumulative hit rate | Hit rate | The cumulative hit rate |
| | 33.2% | 33.2% | 30.1% | 30.1% | 32.9% | 32.9% | | 17.1% | 50.3% | 16.5% | 46.6% | 18.4% | 51.3% | | 18.4% | 68.7% | 11.7% | 58.3% | 14.2% | 65.5% | | 14.1% | 82.8% | 15.6% | 73.9% | 11.3% | 76.8% | | 2.5% | 85.3% | 10.4% | 84.3% | 7.4% | 84.2% | | 3.1% | 88.4% | 3.0% | 87.3% | 7.0% | 91.2% | | 11.6% | 100% | 12.7% | 100% | 8.8% | 100% | Time (ms) | 105032926 (order of magnitude 109) | 1921829 (order of magnitude 107) | 1928397 (order of magnitude 107) |
|
|