A Novel Radial Basis Neural Network-Leveraged Fast Training Method for Identifying Organs in MR Images
Algorithm 1
The Fast-RBF algorithm.
Input: Dataset, approximation parameter , parameter , parameter , parameter , and kernel width , where and
Output: Core-set , Lagrangian coefficient
Training steps:
Step 1: Randomly select 20 samples to form the initial core set ;
Generate the center and radius of the initial CC-MEB according to equation (21) and set the number of iterations t to be 0
Step 2: Randomly select 59 samples and calculate the distance from any sample to the center of the CC-MEB according to equation (22). If there is no sample outside , proceed to step 6
Step 3: Find the farthest sample from the center in step 2 and add the sample to core-set
Step 4: Solve the new CC-MEB, recorded as , and ,
Step 5: Set and return to step 2
Step 6: Terminate the training and return the required output
