A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs
Abstract Human cognition is characterized by a wide range of capabilities including goal-oriented selective attention, distractor suppression, decision making, response inhibition, and working memory. Much research has focused on studying these individual components of cognition in isolation, wherea...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-04-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-32234-y |
| _version_ | 1797845932736249856 |
|---|---|
| author | Sandeep Sathyanandan Nair Vignayanandam Ravindernath Muddapu C. Vigneswaran Pragathi P. Balasubramani Dhakshin S. Ramanathan Jyoti Mishra V. Srinivasa Chakravarthy |
| author_facet | Sandeep Sathyanandan Nair Vignayanandam Ravindernath Muddapu C. Vigneswaran Pragathi P. Balasubramani Dhakshin S. Ramanathan Jyoti Mishra V. Srinivasa Chakravarthy |
| author_sort | Sandeep Sathyanandan Nair |
| collection | DOAJ |
| description | Abstract Human cognition is characterized by a wide range of capabilities including goal-oriented selective attention, distractor suppression, decision making, response inhibition, and working memory. Much research has focused on studying these individual components of cognition in isolation, whereas in several translational applications for cognitive impairment, multiple cognitive functions are altered in a given individual. Hence it is important to study multiple cognitive abilities in the same subject or, in computational terms, model them using a single model. To this end, we propose a unified, reinforcement learning-based agent model comprising of systems for representation, memory, value computation and exploration. We successfully modeled the aforementioned cognitive tasks and show how individual performance can be mapped to model meta-parameters. This model has the potential to serve as a proxy for cognitively impaired conditions, and can be used as a clinical testbench on which therapeutic interventions can be simulated first before delivering to human subjects. |
| first_indexed | 2024-04-09T17:48:00Z |
| format | Article |
| id | doaj.art-392b379b4e414816ac3d4abd638ac091 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-09T17:48:00Z |
| publishDate | 2023-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-392b379b4e414816ac3d4abd638ac0912023-04-16T11:13:00ZengNature PortfolioScientific Reports2045-23222023-04-0113111210.1038/s41598-023-32234-yA generalized reinforcement learning based deep neural network agent model for diverse cognitive constructsSandeep Sathyanandan Nair0Vignayanandam Ravindernath Muddapu1C. Vigneswaran2Pragathi P. Balasubramani3Dhakshin S. Ramanathan4Jyoti Mishra5V. Srinivasa Chakravarthy6Computational Neuroscience Lab, Department of Biotechnology, Bhupat and Jyoti Mehta School of Biosciences, Indian Institute of Technology MadrasComputational Neuroscience Lab, Department of Biotechnology, Bhupat and Jyoti Mehta School of Biosciences, Indian Institute of Technology MadrasComputational Neuroscience Lab, Department of Biotechnology, Bhupat and Jyoti Mehta School of Biosciences, Indian Institute of Technology MadrasNeural Engineering and Translation Labs, Department of Psychiatry, University of California, San DiegoNeural Engineering and Translation Labs, Department of Psychiatry, University of California, San DiegoNeural Engineering and Translation Labs, Department of Psychiatry, University of California, San DiegoComputational Neuroscience Lab, Department of Biotechnology, Bhupat and Jyoti Mehta School of Biosciences, Indian Institute of Technology MadrasAbstract Human cognition is characterized by a wide range of capabilities including goal-oriented selective attention, distractor suppression, decision making, response inhibition, and working memory. Much research has focused on studying these individual components of cognition in isolation, whereas in several translational applications for cognitive impairment, multiple cognitive functions are altered in a given individual. Hence it is important to study multiple cognitive abilities in the same subject or, in computational terms, model them using a single model. To this end, we propose a unified, reinforcement learning-based agent model comprising of systems for representation, memory, value computation and exploration. We successfully modeled the aforementioned cognitive tasks and show how individual performance can be mapped to model meta-parameters. This model has the potential to serve as a proxy for cognitively impaired conditions, and can be used as a clinical testbench on which therapeutic interventions can be simulated first before delivering to human subjects.https://doi.org/10.1038/s41598-023-32234-y |
| spellingShingle | Sandeep Sathyanandan Nair Vignayanandam Ravindernath Muddapu C. Vigneswaran Pragathi P. Balasubramani Dhakshin S. Ramanathan Jyoti Mishra V. Srinivasa Chakravarthy A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs Scientific Reports |
| title | A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs |
| title_full | A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs |
| title_fullStr | A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs |
| title_full_unstemmed | A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs |
| title_short | A generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs |
| title_sort | generalized reinforcement learning based deep neural network agent model for diverse cognitive constructs |
| url | https://doi.org/10.1038/s41598-023-32234-y |
| work_keys_str_mv | AT sandeepsathyanandannair ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT vignayanandamravindernathmuddapu ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT cvigneswaran ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT pragathipbalasubramani ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT dhakshinsramanathan ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT jyotimishra ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT vsrinivasachakravarthy ageneralizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT sandeepsathyanandannair generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT vignayanandamravindernathmuddapu generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT cvigneswaran generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT pragathipbalasubramani generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT dhakshinsramanathan generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT jyotimishra generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs AT vsrinivasachakravarthy generalizedreinforcementlearningbaseddeepneuralnetworkagentmodelfordiversecognitiveconstructs |