A Method to Cluster Long-duration Acoustic Recordings
Yvonne Phillips, Michael Towsey · QUT ePrints (Queensland University of Technology) · 2017
This paper outlines the recording dataset and the method used to choose a clustering algorithm for a large twenty-six month acoustic dataset. The recordings were of the natural environment and consist of thirteen months of recording from each of two sites in two national parks 160 km north of Brisbane, Queensland. This paper also explains the calculation and use of the intra-three-day-distance (I3DD) error measure used to determine the optimum clustering result. Site maps and photos are provided at the end of this document.