Summary: | 碩士 === 國立臺北科技大學 === 自動化科技研究所 === 102 === In the era of technological progress, many large databases to be used to do research, therefore, analysis and processing of information is a very important part, which makes the choice of clustering algorithm becomes very important, and then how to shorten its clustering speed is a question worth considering.
In this thesis, we propose a hardware implementation with a parallel computing capability DBSCAN(Density-based Spatial Clustering of Applications with Noise) clustering algorithm on data clustering. The hardware implementation can search the high-density area for each cluster to classify the data, and remove the noise effectively. In addition, it can classify the irregular data graphics simultaneously. Furthermore, the hardware implementation can reduce the resources complexities substantially and compute the complexities of clustering processing and computing processing. Lastly, the examples are illustrated to show the validity and feasibility of the proposed clustering algorithm and hardware implementation by the simulation result and FPGA hardware architecture.
|