In the research, they analyze the two challenges of learning from time series data with noisy labels: (a) Label noise in time series data misleads the learning process of feature representations, ...

Moscot is scalable to millions of cells, and can track several cell characteristics simultaneously in space and time ... and machine learning enable early, non-perturbative classification of ...