FILES IN THE DISK:
	device_primitives.h : Implementations of inlined device functions, optional
	helper.h : Header file for helper host and device functions
	helper.cu : Implementations of helper host and device functions
	parallel_primitives.h : Header file for parallel primitive algorithms: histogram, transpose, reduction, scan
	parallel_primitives.cu : Implementations of parallel primitive algorithms
	resample_kernels.h : Header file for parallel resampling algorithms
	resample_kernels.cu : Implementations of parallel resampling algorithms
	test_moments > test_moments.cu : Implementation of test for measuring time and MSE of resampling algorithms
	test_moments > plot_results.m : MATLAB script for plotting the results
	test_moments > *.dat : Output files of test_moments program
	test_moments > *.eps : Output files of plot_results.m script
	test_filter > helper.h, helper.cu, parallel_primitives.h, parallel_primitives.cu, filter_kernels.h, filter_kernels.cu : Filter specific implementations and header files of parallel algorithms
	test_filter > test_filter.cu : Implementation of test for measuring time and MSE of particle filters with different resampling algorithms
	test_filter > *.dat : Output files of test_filter program
	test_filter > adaptive > adaptive_plot_results.m : MATLAB script for plotting results for adaptive filter
	test_filter > adaptive > *.eps : Output of adaptive_plot_results.m script
	test_filter > bootstrap > bootstrap_plot_results.m : MATLAB script for plotting results for bootstrap filter
	test_filter > bootstrap > *.eps : Output of bootstrap _plot_results.m script
	smc_em > helper.h, helper.cu, smc_em_resample_kernels.h, smc_em_resample_kernels.cu : SMCEM specific implementations and header files of parallel algorithms
	smc_em > svmodel_smcem.cu : Implementation of test for SMCEM algorithm with bootstrap filter
	smc_em > svmodel_smcem_with_transpose.cu : Implementation of test for SMCEM algorithm with bootstrap filter, butterfly with transpose
	smc_em > svmodel_smcem_less_interaction.cu : Implementation of test for SMCEM algorithm with constrained interaction filter
	smc_em > svmodel_dat_files_* > Output of svmodel_smcem program
	smc_em > svmodel_less_int_dat_files_* > Output of svmodel_smcem_less_int program
	smc_em > plot_results.m : MATLAB script for plotting results for SMCEM with bootstrap filter implementations
	smc_em > plot_results_less_int.m : MATLAB script for plotting results for SMCEM with constrained interaction filter implementations
	smc_em > *.eps : Output of plot_results* scripts
	double_precision > ... : Double precision implementation of speed and MSE tests of resampling algorithms and corresponding results

HARDWARE REQUIREMENTS:
	NVIDIA CUDA capable GPU with compute capability greater than 3.0 and 2 GBs of memory
	4 GBs of main memory
	CPU requirements are not specified for CUDA software, tests are run on a computer with 4 Intel(R) Core(TM) i7-3820 CPU @ 3.60GHz CPU cores

SOFTWARE REQUIREMENTS:
	CUDA Toolkit with version greater than 5.5
	GNU Scientific Library
	MATLAB R2013a or later version
	For Linux systems that support CUDA software see http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-linux/index.html
	For Linux systems GCC compiler and toolchain
	For Windows systems that support CUDA software see http://docs.nvidia.com/cuda/cuda-getting-started-guide-for-microsoft-windows/
	For Windows systems Visual Studio 2010 and Visual C++ 10.0 compiler or later version
