Please enable JavaScript.
Coggle requires JavaScript to display documents.
Cluster A: Parallel and Distributed Clustering Algorithms (DBSCAN…
Cluster A: Parallel and Distributed Clustering Algorithms
BIRCH
PBIRCH
[57]
Scalability, load balancing by cyclic distribution of incoming data
SPMD, message passing share-nothing algorithm scales with number of processors
A-BIRCH
[58]
Parallelized Monte Carlo simulations for each cluster using Apache Spark
DBSCAN
PDSDBSCAN
[60]
Graph algorithm, disjoint-set data structure, bottom-up clusters, better workload, shared & distributed memory
PS-DBSCAN based on Alibaba Cloud
[61]
Fast global union to union disjoint-sets to alleviate communication burden
MR-DBSCAN
[59]
4-stages MapReduce, quick partitioning of non-indexed data, speedup & scaleup
HPDBSCAN
[62]
Computation split heuristic for domain decomposition, data index pre-processing & rule-based cluster merging
Break sequentiality, workload balancing & speed up neighborhood searches
OpenMP/MPI hybrid
Improvement in scale-up, computation time & memory consumption
MR-DBSCAN for heavily skewed data
[64]
Parallelized critical sub-procedures
Novel data partitioning for computational cost estimation
Load balancing, efficiency & scalability for heavily skewed data
G-DBSCAN
[65]
Data indexed using graphs
100x faster than sequential CPU
NG-DBSCAN for arbitrary data
[63]
MapReduce model, arbitrary symmetric distance function
Distributed DBSCAN
[66]
Multinode cluster
High performance, High scalability