Two types of problems

• Neural encoding
• Neural decoding

From Stimulus to Response

• Characterizing stimulus to response difficult
• Neural response is complex, variable
• Spike sequences reflect both intrinsic neural dynamics and temporal characteristics of stimulus
• Spikes may or may not be related to stimulus (multiple connections, feedback loops, etc)
• A concert might be 440Hz, but spikes only go up to 250Hz?
• Neural responses may vary from trial to trial—random processes
• Depends on mood of subject
• Never can predict exact timing of each spike
• Characterizing one neuron → probability
• Multiple neurons → linear algebra
• A spike raster: a clump of spikes that all occur

Spike Trains

• Like impulse train that’s spaced based on a function
• Neural responses can be characterized by firing rates (spikes/sec)
• However, firing rates vary over time—lose time based information by averaging