Dynamic Prediction Research of Silicon Content in Hot Metal Driven by Big Data in Blast Furnace Smelting Process under Hadoop Cloud Platform

Algorithm 3

Parallel design of sample covariance matrix for training set.

//Map

Input: The sample on the current node;

Output: , , .

Main procedure:

1. Scan the current node sample, accumulate the number of samples of the current node ;

2. Calculate ;

3. Calculate ;

//Reduce

Input: Each node , , . ;

Output: .

Main procedure:

1. Summarize the output of the Map node;

2. Find the covariance matrix of the whole sample according to Eq. (26).