Research Article
Applying Data Clustering Feature to Speed Up Ant Colony Optimization
Input parameters: | : Training Set | : The number of classes | : The stop threshold for clustering. | : Initial centroids set. | : A parameter to adjust the size of compact subs-ets . | Output: | (i.e., the set of co-mpact subset, see Figure 1) | , where , and it is comprised by dispersive points | (, see Figure 1) | Void Subroutine 1 () | { | Step 1. Initialization: Let iteration number . Let . Let and , where | denotes empty set. According to initial centroids set , generate initial partition of training set | . | Step 2. While | Step 2.1. Generate new centroids set and new partition | /* Note: Check whether entropy sequence {} is convergent. If it is convergent, | let the convergent marker StableMarker */ | Step 2.2. For | Estimate the entropy of class , that is, . | If | Else | } | /* Note: Extract the data around the centroid of class as a genuine class */ | Step 2.3. For | If | Calculate compact central region according to formula (3) | Calculate : | Let | Let | Update Training Set: | Update centroids set: | | } | } | | } | } |
|