Title: Density Estimation and Clustering
Speaker: Sanjib Sharma, University of Sydney (sanjib.sharma AT gmail.com)
Date: 1pm (AEDT) Wednesday 12th March 2014
Abstract: Estimating density of a given set of points and identifying clusters are two important techniques to reveal hidden information in data. With recent advances in technology, today we have data which is both large (number of objects per data set) and rich (amount of information in each object). This poses a unique challenge for data mining. In multi dimensional spaces, not all the algorithms work equally well. Also, not all algorithms are computationally efficient for analyzing large amounts of data. In this lecture, I will discuss various algorithms and highlight their strengths and weaknesses. I will then concentrate on a few algorithms that work well in multi dimensional spaces and are also fast and efficient to be applied on large data sets. I will also show a few applications of these algorithms to astronomy.
Additional Material: Talk slides are available for download here.