Karpagam Academy of Higher Education

Personalized federated learning for multilingual sentiment analysis poses significant challenges arising from linguistic heterogeneity, non-IID data distributions, and strict privacy requirements. This paper proposes FedPerX, a federated transformer framework that integrates residual adapter-based personalization with adaptive multi-granular differential privacy. The architecture leverages a frozen multilingual backbone (XLM-R) while enabling each client to train lightweight, client-specific adapters. Privacy is enforced through dynamic noise injection at both the feature and adapter levels, calibrated using gradient sensitivity. FedPerX is evaluated on two multilingual benchmarks-MARC and TSMD-spanning structured reviews and informal social media content across more than ten languages. Experimental results demonstrate consistent improvements over seven state-of-the-art baselines, with up to +4.3% gains in macro-F1, a 70% reduction in communication overhead, and the lowest variance in client-level performance. Comprehensive analyses, including fairness, personalization gap, privacy-utility trade-off, and ablation studies, validate the framework's robustness and adaptability. FedPerX advances the design of scalable, personalized, and privacy-preserving models for federated multilingual sentiment analysis.

Abstract - In today’s data-intensive business environment, organizations increasingly depend on Business Intelligence (BI) tools to support evidence-based decision-making and sustain competitive advantage. This study examines how managers at Mind Box LLC perceive and utilize BI tools as Decision Support Systems (DSS) in their routine and strategic decision processes. Using a descriptive research design, primary data were collected from 100 managers across departments through a structured questionnaire, complemented by secondary data from journals and company documents. The analysis employed percentage statistics, chi-square tests, correlation, and one-way ANOVA to evaluate BI usage patterns, perceived benefits, challenges, and satisfaction levels. Results show high adoption of BI tools, with 75% of managers using them at least weekly and a majority reporting moderate to advanced BI knowledge. Managers report that BI tools improve decision quality, reduce uncertainty and risk, enable faster decisions, and contribute to competitive advantage. Key barriers include insufficient training, data quality issues, and system complexity, while no significant differences in satisfaction were found across managerial levels. The findings underscore that BI tools are effective DSS enablers when supported by robust data governance, continuous training, user-friendly design, and a data-driven culture. The study offers practical recommendations for strengthening BI adoption and maximizing its decision support value at Mind Box and similar technology-driven organizations. This study contributes to the growing literature on BI-enabled decision-making in technology-driven organizations. Keywords: Business Intelligence, Decision Support Systems, managerial perception, data-driven decision-making, Mind Box.

Hyperspectral imaging (HSI) is an advanced imaging technique that integrates digital imaging and spectroscopy to acquire detailed spectral information for every pixel in an image, forming a three-dimensional data hypercube. In biomedical research, HSI has enabled a wide range of applications, including real-time tissue and organ analysis during surgical and diagnostic procedures. The technology is capable of operating under challenging conditions such as uncontrolled surgical lighting, tissue motion, and strict real-time decision-making requirements. These platforms support diverse applications, ranging from tissue-level physiological assessment in open and laparoscopic surgeries to cellular-level digital histopathology in dermatology and neurosurgery. Overall, HSI serves as a critical link between clinician expertise and quantitative tissue analysis, demonstrating its transition from research to clinical practice and contributing to improved surgical safety and patient outcomes.

ABSTRACT Improving photovoltaics (PV) system performance through simulation requires accurate PV models. The nonlinear relationship between current and voltage, coupled with incomplete manufacturer data, presents a significant challenge in parameter estimation. This research work presents an innovative optimization framework designed to extract the parameters of a triple‐diode model representing an unknown Solar PV module. The proposed approach leverages a novel hybrid optimization technique called the Kookaburra‐based Dwarf Mongoose (KO‐DM) Optimization Algorithm. It determines the unknown parameter that helps to characterize the equivalent circuit of the PV cell described by the triple‐diode model. The performance of the proposed KO‐DM algorithm was assessed using two distinct datasets: (1) 25°C with 1000 W/m² and (2) 45°C with 1000 W/m². The proposed method offers superior performance, exhibiting the ability to efficiently solve complex nonlinear optimization problems, converge rapidly to the global optimum, and involve a simple computational procedure.

Rare diseases, often called orphan diseases, affect a small number of individuals. They are associated with severe, chronic, debilitating and adverse health outcomes, sometimes leading to death. There are no effective treatments for most of the rare diseases, and hence the patients often rely on therapies that alleviate symptoms, which may require lifetime treatment. This will place a significant burden on patients and their families physically, emotionally and financially, affecting their overall well-being. In India, the healthcare resources are already over-stretched and hence addressing the needs of patients suffering from rare diseases is particularly complex due to the high costs of diagnosis and treatment, lack of awareness, and limited access to specialized services. However, several initiatives have been undertaken at the national level to address the challenges related to the diagnosis, treatment and surveillance of rare diseases. The National Policy for Rare Diseases first introduced in 2017, then revised and launched in 2021, provides a framework for India’s approach in managing rare diseases. Nevertheless, challenges such as inadequate implementation of policy, insufficient manufacturing of orphan drugs at the national level and the lack of an effective surveillance and data collection system hampered the progress in providing an effective public health response for rare diseases. This review provides an overview of the Indian rare disease policy emphasizing government initiatives to tackle the burden of rare diseases on patients, families and the healthcare system.