원문정보
초록
영어
Cluster examination is data mining task for the assignment of collection a set of items in such a path, to the point that questions in the same gathering (called a cluster) are more like one another than to those in different gatherings (clusters). K-means grouping is a technique for group investigation which intends to parcel n perceptions into k groups in which every perception fits in with the cluster with the closest mean. This paper, decided the aftereffect of standard parameter estimations of shading picture division with k-means and the modified k-means with ABC and ACO algorithms. The paper demonstrates that division of color picture with modified k-mean consolidated with swarm Intelligence calculations for color image segmentation gives preferable results over simple k-means and Modified k-means with Ant colony optimization gives better results than modified k-means with Artificial bee colony.
목차
1. Introduction
2. Preliminaries
2.1. K-Means Clustering
2.2. Artificial Bee Colony
2.3. ANT Colony Optimization
3. Proposed Approach
3.1. Modified K-Means
3.2. ABCMK-Means
3.3. ACOMK-Means
4. Implementation and Results
4.1. Accuracy
4.2. Sensitivity
4.3. Specificity
4.4. F-Measure
4.5. Bit Error Rate
4.6. Execution Time
5. Conclusion and Future Work
References