Research Article
Dynamic Prediction Research of Silicon Content in Hot Metal Driven by Big Data in Blast Furnace Smelting Process under Hadoop Cloud Platform
Algorithm 4
Parallel algorithm design of Pegasos’ tth iteration.
//Map | Input: , The number of nodes , the number of random samples taken per iteration ; | Output: The current node obtained , among them: . | Main procedure: | 1. Randomly selected k/M samples; | 2. Define the zero vector ; | 3. For to k/M | If | Then | 4. Solving get | //Reduce | Input: The current number of iterations , , the number of nodes , the number of random samples taken per iteration ; | Output: . | Main procedure: | 1. Calculate ; | 2. Calculate ; | 3. Calculate: | ; | 4. Calculate: | |
|