Latent Dirichlet Allocation (LDA) is a statistical technique for topic modeling. Since it is very computationally demanding, its parallelization has garnered considerable interest. In this paper, we systematically analyze the data access patterns for LDA and devise suitable algorithmic adaptations and parallelization strategies for GPUs. Experiments on large-scale datasets show the effectiveness of the new parallel implementation on GPUs.
CITATION STYLE
Moon, G. E., Nisa, I., Sukumaran-Rajam, A., Bandyopadhyay, B., Parthasarathy, S., & Sadayappan, P. (2018). Parallel Latent Dirichlet Allocation on GPUs. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10861 LNCS, pp. 259–272). Springer Verlag. https://doi.org/10.1007/978-3-319-93701-4_20
Mendeley helps you to discover research relevant for your work.