|Project Description: ||This project is to implement hybrid global cluster algorithms on high-performance
Apache Spark platform. The algorithms combine present cluster algorithms and opti-
mization algorithms for better performance. To be more specific, two kinds of variant
fireflies algorithms are utilized to find global optima among several clustering possi-
bilities to one dataset and use the global optima to initialize k-means algorithm. To
improve the performance of fireflies algorithm, additional features are added into the
algorithm. To test the performance of hybrid improved k-means algorithms, several
experiments are performed on Apache Spark platform based on MapReduce model.