Subekshya Bidari, Department of Applied Mathematics, University of Colorado Boulder
Optimizing flexibility in the collective decisions of honeybees
Honeybees make decisions as a group while searching for a new home site or foraging. The quality of each choice influences the rate at which scout bees recruit others via a waggle dance. In addition, decided bees can influence those with opposing opinions to change their minds via “stop-signals.” Most previous experimental studies have assumed bee swarms make decisions in static environments, but most natural environments are dynamic. In such cases, bees should adapt to new evidence as the environment constantly changes. One way of adapting is to abandon one’s current opinion and restart the evidence-accumulation and decision process (Seeley et al 2012). Incorporating such individual behavior into a dynamical model leads to a collective decision-making process that discounts previous evidence and weights newer information more strongly. We show that properly tuning this “forgetting” process can improve a swarm’s performance on a foraging task in a dynamic environment. Individual forgetfulness allows the group to change its mind, and move to a higher yielding foraging site. Our analysis explores parameter-dependent changes in the foraging yield using bifurcation theory and fast/slow analysis in a mean field version of the collective decision-making model.
Jordan Dixon, Department of Aerospace Engineering Sciences, University of Colorado Boulder
Implementation methods of modeling sensory noise to account for thresholds of human motion perception
The human primarily utilizes vestibular and visual cues to move and orient in one's environment. The brain must combine the network of sensory signals from these peripheral organs to estimate motion of body parts using imperfect information from noisy neurons. Motion thresholds are the physical limits that these systems can detect either self-motion or environmental motion. These thresholds play a fundamental role in balance and locomotion impairment  seen in populations such as elderly or astronauts returning from spaceflight. This study focuses on detection thresholds which represent the ability to correctly determine the direction of translation or rotation. We have tested a preliminary subject group to map vestibular semicircular canal thresholds. An observer model is used to represent the synaptic encoding and neural interpretation, and noise is implemented in attempt to match experimental findings. Various architectures for sensory integration across both vestibular organs (i.e. otoliths and semicircular canals) are proposed, and experimental methods discussed.