Expectation-Maximization Binary Clustering for Behavioural Annotation
Fig 10
C.elegans (Caenorhabditis elegans) search trajectory.
Trivariate clustering of 6 solitary nematode trajectories crawling in an agar plate (with a sampling frequency of 32 Hz and 90 minutes of trajectory time). The clustering is performed at population level (all data points at the same time) and is afterwards visualized on each individual trajectory. The algorithm captures different behaviours in terms of intensity of local search, looping behaviour and relocation.