Rajat Kumar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Rajat Kumar has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Bindu Verma
Delhi Technological University
1 shared publication - Gagan Somashekar
Microsoft, Redmond, Washington, USA
1 shared publication - Helmut H. Strey
State University of New York at Stony Brook
1 shared publication - Lilianne Rivka Mujica-Parodi
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States
1 shared publication - Puneet Sharma
Department of Automation and Process Engineering, Machine Learning Group, UiT – The Arctic University of Norway, Tromsø, Norway
1 shared publication
Latest Publications
- Open Access
- 9 November 2023
- Published by Springer Nature in Computational Brain & Behavior
- Vol. 6 (4) , 556-568
Quantifying Individual Variability in Neural Control Circuit Regulation Using Single-Subject fMRI
As a field, control systems engineering has developed quantitative methods to characterize the regulation of systems or processes, whose functioning is ubiquitous within synthetic systems. In this context, a control circuit is objectively “well regulated” when discrepancy between desired and achieved output trajectories is minimized and “robust” to the degree that it can regulate well in response to a wide range of stimuli. Most psychiatric disorders are assumed to reflect dysregulation of brain circuits. Yet, probing circuit regulation requires fundamentally different analytic strategies than the correlations relied upon for analyses of connectivity and their resultant networks. Here, we demonstrate how well-established methods for system identification in control systems engineering may be applied to functional magnetic resonance imaging (fMRI) data to extract generative computational models of human brain circuits. As required for clinical neurodiagnostics, we show these models to be extractable even at the level of the single subject. Control parameters provide two quantitative measures of direct relevance for psychiatric disorders: a circuit’s sensitivity to external perturbation and its dysregulation.
- 9 November 2023
- Published by Springer Nature in Computational Brain & Behavior
- Vol. 6 (4) , 556-568
Open AccessQuantifying Individual Variability in Neural Control Circuit Regulation Using Single-Subject fMRI
As a field, control systems engineering has developed quantitative methods to characterize the regulation of systems or processes, whose functioning is ubiquitous within synthetic systems. In this context, a control circuit is objectively “well regulated” when discrepancy between desired and achieved output trajectories is minimized and “robust” to the degree that it can regulate well in response to a wide range of stimuli. Most psychiatric disorders are assumed to reflect dysregulation of brain circuits. Yet, probing circuit regulation requires fundamentally different analytic strategies than the correlations relied upon for analyses of connectivity and their resultant networks. Here, we demonstrate how well-established methods for system identification in control systems engineering may be applied to functional magnetic resonance imaging (fMRI) data to extract generative computational models of human brain circuits. As required for clinical neurodiagnostics, we show these models to be extractable even at the level of the single subject. Control parameters provide two quantitative measures of direct relevance for psychiatric disorders: a circuit’s sensitivity to external perturbation and its dysregulation.
- Open Access
- 15 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 39-44
Enhancing the Configuration Tuning Pipeline of Large-Scale Distributed Applications Using Large Language Models (Idea Paper)
- 15 April 2023
- Published by Association for Computing Machinery (ACM)
- p. 39-44
Open AccessEnhancing the Configuration Tuning Pipeline of Large-Scale Distributed Applications Using Large Language Models (Idea Paper)
- 22 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1480-1484
Comparative Analysis of Different hand gestures Detection and Recognition Techniques
Hand Gesture is a wide technology aspect of computer Science. It provides a good technology tool for human computer interaction. With some mathematical algorithms, Humans can interact with computers without using any mechanical devices. Also for deaf and mute people, communication hindrance can be remediate. Hand gesture is the most expressive and frequent among any type of expression language. It can put up sign language recognition to a new level. This paper is oriented towards a comparative analysis of different types of hand gesture Detection and recognition methods. In addition, the result and problems with different approaches are tabulated. These have different gesture types and so different number of gesture classes. Most of these methods produce accuracy of 90-95%. The best accuracy achieved for static hand gestures is 95.7% which is based on the ML algorithm.
- 22 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1480-1484
Comparative Analysis of Different hand gestures Detection and Recognition Techniques
Hand Gesture is a wide technology aspect of computer Science. It provides a good technology tool for human computer interaction. With some mathematical algorithms, Humans can interact with computers without using any mechanical devices. Also for deaf and mute people, communication hindrance can be remediate. Hand gesture is the most expressive and frequent among any type of expression language. It can put up sign language recognition to a new level. This paper is oriented towards a comparative analysis of different types of hand gesture Detection and recognition methods. In addition, the result and problems with different approaches are tabulated. These have different gesture types and so different number of gesture classes. Most of these methods produce accuracy of 90-95%. The best accuracy achieved for static hand gestures is 95.7% which is based on the ML algorithm.