Devdas Gupta
Devdas Gupta has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Anuj Kumar Shah
Yeshiva University, USA
1 shared publication - Bhavna Agrawal
GNIOT Institute of Professional Studies,Commerce Management Department,Greater Noida,INDIA
1 shared publication - E. Muthukumar
Nehru College of Management,Department of MBA,Thirumalayampalayam, Coimbatore
1 shared publication - Ganesh Harke
Independent Researcher, USA
1 shared publication - Lav Kumar1 shared publication
- R. Radhika
B.S Abdur Rahman Crescent Institute of Science and Technology,Department of Commerce,Vandalur Chennai,Tamil Nadu
1 shared publication
Latest Publications
- 5 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 588-594
LLM-Augmented Semantic Resonance System (SRR): A Unified Triple-Objective Architecture for Deep Sequential Recommendation
Sequential recommenders have recently achieved strong gains by modeling the temporal structure of user behavior, yet most systems still rely on shallow item representations that capture co-occurrence statistics rather than the underlying semantics of content. This paper proposes the Semantic Resonance System (SRR), a unified architecture that integrates deep semantic signals from a large language model (LLM) with standard behavioral embeddings for sequential recommendation. SRR uses a domain-adapted Llama 2 13B model, fine-tuned via ParameterEfficient Fine-Tuning (LoRA), to produce high-fidelity semantic item embeddings that are fused with behavioral embeddings through a Dynamic Cross-Attention module. A central component of SRR is the Semantic Resonance Optimization (SRO) loss, which jointly optimizes (i) a Dense All-Action (DAA) loss for high-recall retrieval, (ii) a Binary Cross-Entropy (BCE) loss for calibrated scores, and (iii) a Contrastive Alignment loss $\mathcal{L}_{C A}$ that explicitly aligns semantic and behavioral latent spaces. On the large-scale MovieLens $\mathbf{2 5 M}$ dataset, SRR achieves a $\mathbf{1 6. 4 \%}$ improvement in NDCG@10 over a strong transfer-learning baseline, while also improving diversity, catalog coverage, and cold-start performance. A latency-effectiveness study further shows that SRR can be configured along a smooth Pareto frontier between NDCG@10 and inference cost, supporting deployment under different resource budgets.- 5 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 588-594
LLM-Augmented Semantic Resonance System (SRR): A Unified Triple-Objective Architecture for Deep Sequential Recommendation
Sequential recommenders have recently achieved strong gains by modeling the temporal structure of user behavior, yet most systems still rely on shallow item representations that capture co-occurrence statistics rather than the underlying semantics of content. This paper proposes the Semantic Resonance System (SRR), a unified architecture that integrates deep semantic signals from a large language model (LLM) with standard behavioral embeddings for sequential recommendation. SRR uses a domain-adapted Llama 2 13B model, fine-tuned via ParameterEfficient Fine-Tuning (LoRA), to produce high-fidelity semantic item embeddings that are fused with behavioral embeddings through a Dynamic Cross-Attention module. A central component of SRR is the Semantic Resonance Optimization (SRO) loss, which jointly optimizes (i) a Dense All-Action (DAA) loss for high-recall retrieval, (ii) a Binary Cross-Entropy (BCE) loss for calibrated scores, and (iii) a Contrastive Alignment loss $\mathcal{L}_{C A}$ that explicitly aligns semantic and behavioral latent spaces. On the large-scale MovieLens $\mathbf{2 5 M}$ dataset, SRR achieves a $\mathbf{1 6. 4 \%}$ improvement in NDCG@10 over a strong transfer-learning baseline, while also improving diversity, catalog coverage, and cold-start performance. A latency-effectiveness study further shows that SRR can be configured along a smooth Pareto frontier between NDCG@10 and inference cost, supporting deployment under different resource budgets.- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Explainable AI-Supported Financial Forecasting for EV Charging Infrastructure Investments
The rapid growth of electric vehicles (EVs) has intensified the demand for efficient and strategically planned charging infrastructure, making accurate financial forecasting crucial for sustainable investments. Traditional forecasting approaches often rely on opaque statistical models or historical trends, which limit transparency, hinder investor confidence, and fail to capture complex market dynamics, regulatory changes, and consumer adoption patterns. To address these challenges, this paper proposes an Explainable AI-Supported Financial Forecasting framework for EV Charging Infrastructure Investments (XAI-FFECI), which combines advanced machine learning algorithms with explainable AI techniques to predict revenue streams, operational costs, and investment returns while providing interpretable insights for decision-makers. XAI-FFECI integrates time-series forecasting, feature importance analysis, and scenario-based simulations to capture market volatility, policy interventions, and demand fluctuations. Experimental evaluation conducted using publicly available EV adoption and charging infrastructure datasets demonstrated that XAI-FFECI improved forecasting accuracy by 28%, enhanced interpretability for stakeholders by 34%, and enabled more informed investment decisions compared to conventional black-box financial models. The framework allows investors and policymakers to identify key drivers of profitability, evaluate risk exposure, and optimize resource allocation in EV charging networks. In conclusion, XAI-FFECI provides a scalable, transparent, and adaptive solution for financial planning in EV infrastructure investments, bridging the gap between predictive performance and explainability, and supporting sustainable, datadriven investment strategies in the rapidly evolving electric mobility sector.
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Explainable AI-Supported Financial Forecasting for EV Charging Infrastructure Investments
The rapid growth of electric vehicles (EVs) has intensified the demand for efficient and strategically planned charging infrastructure, making accurate financial forecasting crucial for sustainable investments. Traditional forecasting approaches often rely on opaque statistical models or historical trends, which limit transparency, hinder investor confidence, and fail to capture complex market dynamics, regulatory changes, and consumer adoption patterns. To address these challenges, this paper proposes an Explainable AI-Supported Financial Forecasting framework for EV Charging Infrastructure Investments (XAI-FFECI), which combines advanced machine learning algorithms with explainable AI techniques to predict revenue streams, operational costs, and investment returns while providing interpretable insights for decision-makers. XAI-FFECI integrates time-series forecasting, feature importance analysis, and scenario-based simulations to capture market volatility, policy interventions, and demand fluctuations. Experimental evaluation conducted using publicly available EV adoption and charging infrastructure datasets demonstrated that XAI-FFECI improved forecasting accuracy by 28%, enhanced interpretability for stakeholders by 34%, and enabled more informed investment decisions compared to conventional black-box financial models. The framework allows investors and policymakers to identify key drivers of profitability, evaluate risk exposure, and optimize resource allocation in EV charging networks. In conclusion, XAI-FFECI provides a scalable, transparent, and adaptive solution for financial planning in EV infrastructure investments, bridging the gap between predictive performance and explainability, and supporting sustainable, datadriven investment strategies in the rapidly evolving electric mobility sector.