Incomplete Data Analysis in Smart City Applications

PhD Thesis Proposal Defence


Title: "Incomplete Data Analysis in Smart City Applications"

by

Mr. Siyuan Liu


ABSTRACT:

Incomplete data in my work is defined as the data with extremely limited 
samples observed, which brings big challenges to data mining. Such extremely 
limited sample data obviously gives us terrible bias and inaccurate results. 
Given one over ten thousand of the whole set of vehicles in a city, how can we 
still retrieve the vehicle distribution and detect the hot spots/crowded areas 
in the city? The traditional density-based clustering methods work not well 
because of the very limited and errorable vehicle density/location information. 
Hence we need new algorithms to handle such incomplete data, in terms of 
accuracy and scalability. On the other hand, the vehicle traces are typical 
spatio-temporal data, which requires efficient approaches. In this paper, we 
have an interesting observation that the vehicle speed can indicate the 
crowdedness of a given area. In other words, if a given area is very crowded, 
then the vehicles’ speed in this area is low; while if this area is not 
crowded, then the vehicles’ speed in this area prefers high. As such the 
mobility of samples is naturally incorporated and a novel non-density-based 
clustering method is developed, called mobility-based clustering. Several key 
factors beyond the vehicle crowdedness have been identified and techniques to 
compensate these effects are proposed. We evaluate the performance of 
mobility-based clustering based on real traffic situations. Experimental 
results show that using 0.3 % of vehicles as the samples, mobility-based 
clustering can accurately identify hot spots which can hardly be obtained by 
the latest representative algorithm UMicro.


Date:                   Tuesday, 28 June 2011

Time:                   3:30pm - 5:30pm

Venue:                  Room 4483
                         lifts 25/26

Committee Members:      Prof. Lionel Ni (Supervisor)
                         Prof. Shing-Chi Cheung (Chairperson)
 			Dr. Qiong Luo
 			Dr. Raymond Wong


**** ALL are Welcome ****