An Agent-Based Computational Framework for Distributed Data Analysis

Fukuda, Munehiro; Gordon, Collin; Mert, Utku; Sell, Matthew

doi:10.1109/mc.2019.2932964

Cited by 7 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 3 compares these three libraries in LoC, focusing on only their logics. This is because our previous work [31,40] has confirmed that LoC of MapReduce is always larger than the other two libraries, due to its considerable percentage of boilerplate LoC that corresponds to parallelization but non-logic code. Except ESP, Spark demonstrates its smallest LoC.…”

Section: Programmability In Parallelizationmentioning

confidence: 86%

“…The ESP benchmark (see Figure 8) compares the MASS library's entire execution with MapReduce and Spark's trapezoidal decomposition. This is because, from our previous work [31,40], MapReduce and Spark' BFS execution only gives additional overheads, (i.e., 17-27 seconds with 3,000 vertices). The performance measurements show the MASS library's CPU scalability with up to two and four computing nodes for 20K and 40K datasets respectively.…”

Section: Parallel Performancementioning

confidence: 93%

“…3. We have already demonstrated agents' computational superiority to data streaming in some geometric and graph problems such as nearest neighbors, minimum spanning trees, and BFS [34,40], and thus we exclude them from this paper's scope. shows all points residing in a given space.…”

Section: Parallelized Algorithmsmentioning

confidence: 99%

See 2 more Smart Citations

Agent-Based versus Data-Streamed Big-Data Parallelization in Computational Geometry

Fukuda,

Potturi,

Helzerman

2023

Preprint

Self Cite

View full text Add to dashboard Cite

While some spatial analytics in GIS are well-supported by combinations of data-streaming tools and machine-learning algorithms, (e.g., Spark and KNN on a geographic map), the others are tightly coupled with computational geometry. They handle multi-dimensional or graph structures that should be better maintained in distributed memory for the purpose of dynamic spatial analysis and repetitive geometric queries. We expect that dispatching agents as active data analyzers into datasets would work smoother than disassembling and conveying datasets into passive data-streaming tools. This research compared agent-based and data-streaming big-data parallelization in computational geometry. We parallelized the closest pair of points, the convex hull, the Euclidean shortest path, the point location, and the range search problems, using our MASS (multi-agent spatial simulation) library, MapReduce, and Spark. MASS performed best in three of the programs, which demonstrated agents’ strength in spatial analysis over many data points and repetitive geometric queries with deep tree traverses.

show abstract

Section: Programmability In Parallelizationmentioning

confidence: 86%

Section: Parallel Performancementioning

confidence: 93%