Skip to content

Fast and Scalable MapReduce-based Vertical Mining

When:
Thursday, 12th July 2018 1:30 pm
Where: EITC E2-528

Speaker: Jialiang Yu

Abstract:

Mining uncertain data is challenging because uncertainty is usually represented as floating-point/real numbers which are infinite (cf. representing finite occurrence counts when mining precise data). This means that they are not easy to store in a data structure. Although there exist some data mining algorithms for handling uncertain data, these algorithms become inefficient when size of data becomes so big. Vertical data mining algorithms have advantages in that they run fast and require low memory space. Hence, for my MSc thesis, I propose two vertical mining algorithms that mine big uncertain data. Analytical and experimental evaluation results show that, between the two algorithms, MR-UV-Eclat is fast and scalable.

Complete Seminar List
© 2011 University of Manitoba Department of Computer Science
Back to top