Parallel Gaussian Process with Kernel Approximation in CUDA Implementation and Benchmarking

Implementing parallel Gaussian process with kernel approximation in CUDA to improve performance.