Measuring Fisher information accurately in correlated neural populations.
Neural responses are known to be variable. In order to understand how this neural variability constrains behavioral performance, we need to be able to measure the reliability with which a sensory stimulus is encoded in a given population. However, such measures are challenging for two reasons: First...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2015-06-01
|
| Series: | PLoS Computational Biology |
| Online Access: | http://europepmc.org/articles/PMC4451760?pdf=render |
| _version_ | 1811263811198386176 |
|---|---|
| author | Ingmar Kanitscheider Ruben Coen-Cagli Adam Kohn Alexandre Pouget |
| author_facet | Ingmar Kanitscheider Ruben Coen-Cagli Adam Kohn Alexandre Pouget |
| author_sort | Ingmar Kanitscheider |
| collection | DOAJ |
| description | Neural responses are known to be variable. In order to understand how this neural variability constrains behavioral performance, we need to be able to measure the reliability with which a sensory stimulus is encoded in a given population. However, such measures are challenging for two reasons: First, they must take into account noise correlations which can have a large influence on reliability. Second, they need to be as efficient as possible, since the number of trials available in a set of neural recording is usually limited by experimental constraints. Traditionally, cross-validated decoding has been used as a reliability measure, but it only provides a lower bound on reliability and underestimates reliability substantially in small datasets. We show that, if the number of trials per condition is larger than the number of neurons, there is an alternative, direct estimate of reliability which consistently leads to smaller errors and is much faster to compute. The superior performance of the direct estimator is evident both for simulated data and for neuronal population recordings from macaque primary visual cortex. Furthermore we propose generalizations of the direct estimator which measure changes in stimulus encoding across conditions and the impact of correlations on encoding and decoding, typically denoted by Ishuffle and Idiag respectively. |
| first_indexed | 2024-04-12T19:53:16Z |
| format | Article |
| id | doaj.art-6430ddabc2ef475190291a77f144c8c7 |
| institution | Directory Open Access Journal |
| issn | 1553-734X 1553-7358 |
| language | English |
| last_indexed | 2024-04-12T19:53:16Z |
| publishDate | 2015-06-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS Computational Biology |
| spelling | doaj.art-6430ddabc2ef475190291a77f144c8c72022-12-22T03:18:46ZengPublic Library of Science (PLoS)PLoS Computational Biology1553-734X1553-73582015-06-01116e100421810.1371/journal.pcbi.1004218Measuring Fisher information accurately in correlated neural populations.Ingmar KanitscheiderRuben Coen-CagliAdam KohnAlexandre PougetNeural responses are known to be variable. In order to understand how this neural variability constrains behavioral performance, we need to be able to measure the reliability with which a sensory stimulus is encoded in a given population. However, such measures are challenging for two reasons: First, they must take into account noise correlations which can have a large influence on reliability. Second, they need to be as efficient as possible, since the number of trials available in a set of neural recording is usually limited by experimental constraints. Traditionally, cross-validated decoding has been used as a reliability measure, but it only provides a lower bound on reliability and underestimates reliability substantially in small datasets. We show that, if the number of trials per condition is larger than the number of neurons, there is an alternative, direct estimate of reliability which consistently leads to smaller errors and is much faster to compute. The superior performance of the direct estimator is evident both for simulated data and for neuronal population recordings from macaque primary visual cortex. Furthermore we propose generalizations of the direct estimator which measure changes in stimulus encoding across conditions and the impact of correlations on encoding and decoding, typically denoted by Ishuffle and Idiag respectively.http://europepmc.org/articles/PMC4451760?pdf=render |
| spellingShingle | Ingmar Kanitscheider Ruben Coen-Cagli Adam Kohn Alexandre Pouget Measuring Fisher information accurately in correlated neural populations. PLoS Computational Biology |
| title | Measuring Fisher information accurately in correlated neural populations. |
| title_full | Measuring Fisher information accurately in correlated neural populations. |
| title_fullStr | Measuring Fisher information accurately in correlated neural populations. |
| title_full_unstemmed | Measuring Fisher information accurately in correlated neural populations. |
| title_short | Measuring Fisher information accurately in correlated neural populations. |
| title_sort | measuring fisher information accurately in correlated neural populations |
| url | http://europepmc.org/articles/PMC4451760?pdf=render |
| work_keys_str_mv | AT ingmarkanitscheider measuringfisherinformationaccuratelyincorrelatedneuralpopulations AT rubencoencagli measuringfisherinformationaccuratelyincorrelatedneuralpopulations AT adamkohn measuringfisherinformationaccuratelyincorrelatedneuralpopulations AT alexandrepouget measuringfisherinformationaccuratelyincorrelatedneuralpopulations |