@article{thomas_topological_2021, abstract = {Video of nematodes/roundworms was analyzed using persistent homology to study locomotion and behavior. In each frame, an organism's body posture was represented by a high-dimensional vector. By concatenating points in fixed-duration segments of this time series, we created a sliding window embedding (sometimes called a time delay embedding) where each point corresponds to a sequence of postures of an organism. Persistent homology on the points in this time series detected behaviors and comparisons of these persistent homology computations detected variation in their corresponding behaviors. We used average persistence landscapes and machine learning techniques to study changes in locomotion and behavior in varying environments.}, author = {Thomas, Ashleigh and Bates, Kathleen and Elchesen, Alex and Hartsock, Iryna and Lu, Hang and Bubenik, Peter}, date = {2021-02-18}, eprint = {2102.09380}, eprinttype = {arxiv}, journaltitle = {{arXiv}:2102.09380 [math, q-bio]}, keywords = {1 - C. Elegans, 1 - Neuroscience, 1 - Quantitative Biology, 2 - Machine learning, 2 - Persistence landscape, 2 - Persistent homology, 2 - Sliding windows, 2 - Time series, 3 - 2D images, 3 - Video}, title = {Topological data analysis of C. elegans locomotion and behavior}, url = {http://arxiv.org/abs/2102.09380}, urldate = {2021-02-19} }