SURVEY ON BIG DATA DIMENSIONALITY REDUCTION

  • Unique Paper ID: 158820
  • Volume: 9
  • Issue: 10
  • PageNo: 888-894
  • Abstract:
  • Dimensionality reduction is a common problem in scientific simulations, particularly in the field of transientsimulations, where the number of variables can be very large. One approach to addressing this problem is to usediffusion maps, a nonlinear dimensionality reduction method that is based on the concept of diffusion on a graph. Intransientsimulations,thegoalisoftentounderstandtheevolutionofasystemovertime.Thiscanbechallengingwhenthe system has a high number of variables, as it can be difficult to visualize and analyze the data. Dimensionalityreduction techniques can be used for many numbers of variable detection, making it easier to analyze and understandthedata. Diffusion maps is a nonlinear dimensionality method coming from the idea of diffusion on a graph. It uses theeigenvectors of the graph Laplacian to identify useful andneeded variablesin the system and to project the givendata onto a dimensional space which is lower . This allows forvisualization and also analysis of the data in a moremanageable form. In the context of transient simulations, diffusion maps can be used to identify the most importantvariables and to track the evolution of the system over time. It can also be used to identify patterns and relationshipsin the data that may not be evident in the original, high-dimensional space. Overall, the use of diffusion maps intransientsimulationscanprovidevaluableinsightsintothebehaviorofthesystemandcanfacilitatetheunderstandingandanalysisof complex, high-dimensionaldata.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 9
  • Issue: 10
  • PageNo: 888-894

SURVEY ON BIG DATA DIMENSIONALITY REDUCTION

Related Articles