Accelerating the Critical Line Algorithm for Portfolio Optimization Using GPUs
Department of Computer Science and Engineering, University of Nevada, Reno
2016 International Conference on Information Technology: New Generations (ITNG), 2016
@article{singh2016accelerating,
title={Accelerating the Critical Line Algorithm for Portfolio Optimization Using GPUs},
author={Singh, Raja H and Barford, Lee and Harris Jr, Frederick},
year={2016}
}
Since the introduction of the Modern Portfolio Theory by Markowitz in the Journal of Finance in 1952, it has been the underlying theory in several portfolio optimization techniques. With the advancement of computers, most portfolio optimization are done by CPUs. Over the years, there have been papers that introduce various optimization methods including those introduced by Markowitz, implemented on the CPUs. In the recent years, GPUs have taken the front seat as a technology to take computational speeds to new levels. However, very few papers published about portfolio optimization utilize GPUs. Our paper is about accelerating the open source Critical Line Algorithm for portfolio optimization by using GPU’s, more precicely using NVIDIA’s GPUS and CUDA API, for time consuming parts of the algorithm.
February 1, 2016 by hgpu
