In this talk, I demonstrate a real world application of using Alien::OpenMP, Inline::C, OpenMP::Environment, C, and OpenMP to implement a threaded version of the DBSCAN (density-based spatial clustering of applications with noise) algorithm.
This algorithm is typically used as an alternative to k-clustering in machine learning and data mining/knowledge discover when the number of clusters is not well understood ahead of time.
The dataset in this case represents a physical domain, but is clustered 2-dimensionally in away that make this approach potentially useful. I will introduce the dataset sufficiently to provide the background and motivation.
The talk will include any updates about the Perl+OpenMP project hosted on Github and provide a few well earned tips for taking advantage of C/OpenMP in your Perl.
Create a custom schedule.
Take it with you on mobile.
Get listed in the directory.
