Spike Train decoding

1
(No Transcript)
2
Spike Train decoding
3
Spike Train decoding
4
Spike Train decoding
5
Summary

• Decoding of stimulus from response
• Two choice case
• Discrimination
• ROC curves
• Population decoding
• MAP and ML estimators
• Bias and variance
• Fisher information, Cramer-Rao bound
• Spike train decoding

6
Chapter 4
7
Entropy
8
Entropy
9
Mutual information
10
Mutual information
11
KL divergence
12
Continuous variables
13
Entropy maximization
14
Entropy maximization
15
Population of neurons
16
Retinal Ganglion Cell Receptive Fields
17
Retinal Ganglion Cell Receptive Fields
18
Retinal Ganglion Cell Receptive Fields
19
Retinal Ganglion Cell Receptive Fields
20
Retinal Ganglion Cell Receptive Fields
21
Temporal processing in LGN
22
Temporal processing in LGN
23
Temporal processing in LGN
24
Temporal vs spatial coding
25
Entropy of spike trains
26
Entropy of spike trains
27
Entropy of spike trains
28
Entropy of spike trains

• Spike train mutual information measurements
  quantify stimulus specific aspects of neural
  encoding.
• Mutual information of bullfrog peripheral
  auditory neurons was estimated
• 1.4 bits/sec for broadband noise stimulus
• 7.8 bits/sec for bullfrog call-like stimulus

29
Summary

• Information theory quantifies how much a response
  says about a stimulus
• Stimulus, response entropy
• Noise entropy
• Mutual information, KL divergence
• Maximizing information transfer yields biological
  receptive fields
• Factorial codes
• Equalization
• Whitening
• Spike train mutual information