GPU-Based Regression Analysis on Sparse Grids
نویسنده
چکیده
Prediction and forecasting has become very important in modern society. Regression analysis enables to predict easily based on given data. This paper focuses on regression analysis on sparse grids using the existing toolbox Sparse Grid ++ (SG). The core workload of the regression analysis will be implemented on graphics cards using NVIDIA’s Compute Unified Device Architecture (CUDA). Therefore, we give guidance how to get high performance when dealing with this particular problem using CUDA enabled graphics cards. We also focus on problems where the datasets are larger than the available device memory. Finally, we present test results for real-world and artificial datasets.
منابع مشابه
Fast Fluid Simulations with Sparse Volumes on the GPU
We introduce efficient, large scale fluid simulation on GPU hardware using the fluid-implicit particle (FLIP) method over a sparse hierarchy of grids represented in NVIDIA R © GVDB Voxels. Our approach handles tens of millions of particles within a virtually unbounded simulation domain. We describe novel techniques for parallel sparse grid hierarchy construction and fast incremental updates on ...
متن کاملGPU-based Dynamic Tubular Grids for Sparse Volume Rendering
Dynamic Tubular Grids (DT-Grids) are designed to encode gridaligned data in level-set simulations. While they are extremely efficient for storing sparse volumetric data, they require logarithmic time for random access. We demonstrate that DT-Grids can be used to efficiently render sparse volumetric data that would otherwise not be able to fit in texture memory. For many real-world biomedical da...
متن کاملAlgorithmes sur GPU de visualisation et de calcul pour des maillages non-structurés
Most recent algorithms for Geometry Processing or Computational Fluid Dynamics (CFD) areusing new types of grids made of arbitrary polyhedra, in other words strongly unstructured grids. Incase of CFD simulations, these grids can be mapped with scalar or vector fields representing physicalproperties (for example : density, porosity, permeability).This thesis proposes new tools fo...
متن کاملEfficient Regression for Big Data Problems using Adaptive Sparse Grids
The amount of available data increases rapidly. This trend, often related to as Big Data challenges modern data mining algorithms, requiring new methods that can cope with very large, multi-variate regression problems. A promising approach that can tackle non-linear, higher-dimensional problems is regression using sparse grids. Sparse grids use a multiscale system of grids with basis functions ...
متن کاملTime Series Forecasting using Sparse Grids
We present a machine learning approach for the forecasting of time series using the sparse grid combination technique. In this approach, the problem of analyzing a time series is first transformed into a higher-dimensional regression problem based on a delay embedding of the empirical data. Then, a grid-based approach is used to discretize the resulting high-dimensional feature space. In order ...
متن کامل