File(s) stored somewhere else
Please note: Linked content is NOT stored on Carnegie Mellon University and we can't guarantee its availability, quality, security or accept any liability.
Efficient Temporal Pattern Mining for Humanoid Robot
Any type of content formally published in an academic journal, usually following a peer-review process.
Pattern mining in temporal databases is one of the challenging platform which holds attention when some ordered sequences are frequently occurred at different time instances in the dataset. We have found temporal patterns in humanoid robot dataset of HOAP-2 (Humanoid Open-Architecture Platform) which generates different motions through recurring sequences of various joint associations. For mining temporal patterns in that dataset we have proposed a method. This method uses FP-Temporal and SH(Soft-Hyperlinked)-Temporal mining algorithm as pattern growth methods for generating temporal association rules for various motion patterns of HOAP-2. Brief performance analysis shows that SH-Temporal is much efficient than FP-Temporal for such datasets and works significantly for mining sequentially associative temporal patterns in terms of temporal association rules.