Title: Functional Data Analysis to Describe and Classify Southern Resident Killer Whale Calls
Speaker: Paul Nguyen Hong Duc (杏吧原创 University)
Date:聽 Friday, February 3rd, 2023
Time:聽 1:00 鈥� 2:00 p.m.
Location: HP 4351 (Macphail Room)
础产蝉迟谤补肠迟:听 The Southern Resident killer whale (SRKW) is an聽endangered聽聽population of whales found in聽the northeast Pacific. They have a vocal dialect unique from other killer聽whales, having a repertoire聽of distinct stereotyped calls. A framework for聽distinguishing SRKW call types using the frequency traces of the amplitude聽ridges from their spectrograms (termed frequency ridges) is proposed.聽The first聽step is the extraction of these frequency ridges of SRKW calls using an聽Sequential Monte Carlo (SMC) approach. Next, these frequency ridges are聽converted into functional data using聽B-spline functions. They are analysed with聽a functional principal component (FPC) analysis to characterise the intrinsic聽variability of frequency ridges within a call type. The FPCs are able to聽capture the general patterns in the frequency ridges of the different SRKW call聽types. The FPCs are also used as the basis for call classification. This聽framework proves to be successful for聽classification with some call types聽correctly classified almost 80\% of the time, while other calls are less well聽discriminated. On balance, this approach showed reasonable performance given聽the small sample size available.