Anuj Sirohi | Jawaharlal Nehru University (original) (raw)
Uploads
Papers by Anuj Sirohi
Proceedings of the ... AAAI Conference on Artificial Intelligence, Mar 24, 2024
Ensuring fairness in Recommendation Systems (RSs) across demographic groups is critical due to th... more Ensuring fairness in Recommendation Systems (RSs) across demographic groups is critical due to the increased integration of RSs in applications such as personalized healthcare, finance, and e-commerce. Graph-based RSs play a crucial role in capturing intricate higher-order interactions among entities. However, integrating these graph models into the Federated Learning (FL) paradigm with fairness constraints poses formidable challenges as this requires access to the entire interaction graph and sensitive user information (such as gender, age, etc.) at the central server. This paper addresses the pervasive issue of inherent bias within RSs for different demographic groups without compromising the privacy of sensitive user attributes in FL environment with the graph-based model. To address the group bias, we propose F 2 PGNN (Fair Federated Personalized Graph Neural Network), a novel framework that leverages the power of Personalized Graph Neural Network (GNN) coupled with fairness considerations. Additionally, we use differential privacy techniques to fortify privacy protection. Experimental evaluation on three publicly available datasets showcases the efficacy of F 2 PGNN in mitigating group unfairness by 47% ∼ 99% compared to the state-of-the-art while preserving privacy and maintaining the utility. The results validate the significance of our framework in achieving equitable and personalized recommendations using GNN within the FL landscape.
arXiv (Cornell University), Feb 20, 2024
arXiv (Cornell University), Dec 9, 2023
Journal of Physics: Conference Series, 2015
Many existing studies on pattern formation in the reaction-diffusion systems rely on deterministi... more Many existing studies on pattern formation in the reaction-diffusion systems rely on deterministic models. However, environmental noise is often a major factor which leads to significant changes in the spatiotemporal dynamics. In this paper, we focus on the spatiotemporal patterns produced by the predator-prey model with ratio-dependent functional response and density dependent death rate of predator. We get the reaction-diffusion equations incorporating the self-diffusion terms, corresponding to random movement of the individuals within two dimensional habitats, into the growth equations for the prey and predator population. In order to have to have the noise added model, small amplitude heterogeneous perturbations to the linear intrinsic growth rates are introduced using uncorrelated Gaussian white noise terms. For the noise added system, we then observe spatial patterns for the parameter values lying outside the Turing instability region. With thorough numerical simulations we characterize the patterns corresponding to Turing and Turing-Hopf domain and study their dependence on different system parameters like noise-intensity, etc.
The IT Industry in India has grown with an exceptionally high growth rate in the post-reform year... more The IT Industry in India has grown with an exceptionally high growth rate in the post-reform years and contributed a large share to the national GDP. Despite the uncertain global economic scenario, the IT industry has steadily augmented and accelerated the growth of India. This Industry absorbs a large pool of Indian skilled human resources which makes the country a global IT hub. The IT Industry has been instrumental in transforming the whole Indian economic and governance landscape. India’s IT industry is gaining footsteps in new disruptive technologies and will play a leading role in the ongoing fourth industrial revolution globally
Proceedings of the ... AAAI Conference on Artificial Intelligence, Mar 24, 2024
Ensuring fairness in Recommendation Systems (RSs) across demographic groups is critical due to th... more Ensuring fairness in Recommendation Systems (RSs) across demographic groups is critical due to the increased integration of RSs in applications such as personalized healthcare, finance, and e-commerce. Graph-based RSs play a crucial role in capturing intricate higher-order interactions among entities. However, integrating these graph models into the Federated Learning (FL) paradigm with fairness constraints poses formidable challenges as this requires access to the entire interaction graph and sensitive user information (such as gender, age, etc.) at the central server. This paper addresses the pervasive issue of inherent bias within RSs for different demographic groups without compromising the privacy of sensitive user attributes in FL environment with the graph-based model. To address the group bias, we propose F 2 PGNN (Fair Federated Personalized Graph Neural Network), a novel framework that leverages the power of Personalized Graph Neural Network (GNN) coupled with fairness considerations. Additionally, we use differential privacy techniques to fortify privacy protection. Experimental evaluation on three publicly available datasets showcases the efficacy of F 2 PGNN in mitigating group unfairness by 47% ∼ 99% compared to the state-of-the-art while preserving privacy and maintaining the utility. The results validate the significance of our framework in achieving equitable and personalized recommendations using GNN within the FL landscape.
arXiv (Cornell University), Feb 20, 2024
arXiv (Cornell University), Dec 9, 2023
Journal of Physics: Conference Series, 2015
Many existing studies on pattern formation in the reaction-diffusion systems rely on deterministi... more Many existing studies on pattern formation in the reaction-diffusion systems rely on deterministic models. However, environmental noise is often a major factor which leads to significant changes in the spatiotemporal dynamics. In this paper, we focus on the spatiotemporal patterns produced by the predator-prey model with ratio-dependent functional response and density dependent death rate of predator. We get the reaction-diffusion equations incorporating the self-diffusion terms, corresponding to random movement of the individuals within two dimensional habitats, into the growth equations for the prey and predator population. In order to have to have the noise added model, small amplitude heterogeneous perturbations to the linear intrinsic growth rates are introduced using uncorrelated Gaussian white noise terms. For the noise added system, we then observe spatial patterns for the parameter values lying outside the Turing instability region. With thorough numerical simulations we characterize the patterns corresponding to Turing and Turing-Hopf domain and study their dependence on different system parameters like noise-intensity, etc.
The IT Industry in India has grown with an exceptionally high growth rate in the post-reform year... more The IT Industry in India has grown with an exceptionally high growth rate in the post-reform years and contributed a large share to the national GDP. Despite the uncertain global economic scenario, the IT industry has steadily augmented and accelerated the growth of India. This Industry absorbs a large pool of Indian skilled human resources which makes the country a global IT hub. The IT Industry has been instrumental in transforming the whole Indian economic and governance landscape. India’s IT industry is gaining footsteps in new disruptive technologies and will play a leading role in the ongoing fourth industrial revolution globally