Manifesto
The dramatic increase in availability of computational resources in recent years affords neuroscientists an unprecedented opportunity to employ the rich array of quantitative concepts developed in the fields of communications/networking, information theory, nonlinear circuits, signal processing and machine learning to address outstanding questions regarding information representation and processing in biological sensory systems.
In order to effectively leverage these concepts to advance our understanding of the functional principles underlying sensory systems, researchers in the biological sciences must be prepared to employ paradigms from the world of engineering.
To wit, we believe that formal methods of time-domain encoding and computation should serve as a theoretical basis for understanding information representation and processing in neural systems. This approach encourages the development of models of neural computation that not only account for the observed behavior of isolated neurons or neural microcircuits, but provide reusable constructs that may be ultimately employed in elucidating the processing performed by multineuron networks of high organizational complexity.
The Bionet Group is supported by grants from
 |
 |