1
Lecture 52 Computational Neuroethology Scientific Method
Observe natural phenomenon. Summarize observations in the form of a working or tentative hypothesis that is consistent with the
- bservations.
Use the hypothesis to generate predictions.
MODEL
Use the hypothesis to generate predictions. Test predictions by performing experiments with proper controls or by making additional observations. Accept or reject hypothesis. if accepted: develop hypothesis further, or make additional observations. if rejected: modify hypothesis, or consider an alternative hypothesis that is also consistent with original
- bservations and with results of experiments.
Models and Fundamental Questions in Neuroscience and in Neuroethology
How does the brain…
- process sensory information?
- recognize stimuli?
- make decisions?
- coordinate motor acts?
- remember previous stimuli experienced or? motor acts
remember previous stimuli experienced or? motor acts performed?
ANALYSIS LEVEL
whole animal molecular subcellular cellular circuits
Forms of Models
- Equations describing dynamic processes.
– Solutions to differential equations by analytic methods
(rare)
– Computer simulation by iteration (more common).
- Neural networks (actually a system of interconnected
elements each governed by equations describing synaptic elements each governed by equations describing synaptic inputs, weights, thresholds and outputs)
– solutions by neuromimes (electronic mimics of elements) – or by computer solution to matrix algebra
- Electronic simulators of circuit: chips that mimic circuits in
the brain.
- Robots: working models in silicon and metal
Levels of Modeling
- Space clamped axon model. Data from v-clamp
- experiments. Hodgkin & Huxley models simulate
the form of action potentials.
- Cable theory plus HH: importance of cell
geometry, morphology of dendrites. g y, p gy
- More than Na+ and K+ channels: fine tuning the
spike.
- Analysis of small networks.
- Modeling the Brain.