Learning run-and-tumble chemotaxis with support vector machines

Abstract To navigate in spatial fields of sensory cues, bacterial cells employ gradient sensing by temporal comparison for run-and-tumble chemotaxis. Sensing and motility noise imply trade-off choices between precision accuracy. gain insight into these trade-offs, we learn optimal chemotactic decision filters using supervised machine learning, applying support vector machines to a biologically motivated training dataset. We discuss how the filter depends on level noise, derive an empirical power law measurement time with $\alpha =0.2, \ldots ,0.3$ as function rotational diffusion coefficient D rot characterizing noise. A weak amount slightly increases performance.