Year 2025 | Volume - 2 | Issue - 1 [January - June 2025]

Dynamic Reduction Factors in SEIR Models: Enhancing Disease Dynamics Prediction

Author Name: Santosh C J and Dr. Anurag Shakya

Paper id: 25201

Abstract:

The SEIR model is a foundational framework in epidemiology used to analyze the transmission of infectious diseases. However, traditional models often treat susceptibility as static, neglecting the real-time effects of interventions such as vaccination campaigns and increased exposure awareness. This research introduces dynamic susceptibility reduction factors into the SEIR model to better reflect evolving disease dynamics. By modifying the compartmental equations, we account for changes in the susceptible population over time, influenced by exposure rates and immunization efforts. Using hypothetical data and standard epidemiological parameters, we demonstrate the enhanced model’s behaviour over a 10-day simulation. Results indicate that dynamic reduction factors significantly slow the spread of disease and provide a more realistic depiction of outbreak progression. This improvement enhances the model’s predictive power and supports more effective public health decision-making. Incorporating real-time factors into disease modelling is essential for accurate forecasting and for guiding timely interventions during epidemic or pandemic events.

Adaptive Memetic Algorithm for Solving Complex Graph Coloring Problem

Author Name: Anjali Gangrade, Bhawna Agrawal and Sanjit Kumar

Paper id: 25202

Abstract:

Graph vertex coloring with a fixed number of colors is a well-known, extensively studied NP-complete problem. The most effective approaches to solving this challenge have been hybrid algorithms, such as memetic algorithms or quantum annealing, which combine the strengths of local search within population-based frameworks. In this paper, we address a production scheduling problem involving three distinct car models-XUV, Sedan, and Mini XUV requiring assignment to two painting lines and two assembly lines. The objective is to minimize total production time while adhering to various operational constraints such as model compatibility for painting lines, specific assembly line restrictions, and differing painting and assembly times for each model. To solve this problem efficiently, we propose a solution based on the Memetic Graph Colouring Algorithm. Our approach incorporates a graph-based representation of the problem, where nodes represent tasks (painting and assembly of individual car units) and edges represent incompatibilities or resource conflicts. The Memetic Algorithm optimizes the assignment of tasks to time slots, leveraging both crossover and mutation operations to explore feasible solutions. Specifically, mutation is employed to randomly alter schedules by swapping tasks within slots, which introduces diversity and enhances the search process.

Timestamp-Based Validation Approach for Detecting Stale Data in Asynchronous Microservice Architecture

Author Name: Sandeep Sharma and Vijay Pal Singh

Paper id: 25203

Abstract:

Data freshness is a crucial problem for microservice architectures as services are decoupled and produce the effects of updates in an asynchronous manner. In these contexts, addressing the issue of stale data across service boundaries is still a pressing issue. This paper introduces the Timestamp-Based Validation Approach (TBVA), a lightweight method that can be employed to identify stale information without the overhead of strong consistency models or a central coordinating process. With TBVA, services can evaluate data freshness at query time by referring to short-lived temporal metadata updated with recent update activity. This metadata is kept outside in a cache store like Redis, administrated by expiration policies to balance performance and reliability. Services detect and mark potentially stale data by comparing the cache timestamp with the record’s modification timestamp in the database before returning responses. The technique increases the integrity of cross-service data and allows responsiveness and architectural freedom. TBVA is particularly suitable for event-driven, asynchronous microservice systems, in which real-time correct processing is critical while achieving global transactional consistency is also not feasible.

Phytochemical analysis and secondary metabolite characterization of different extracts of Citrus fruits peel

Author Name: Ranjit Kumar, Dr Saurabh Mishra, Dr Vijay Nema

Paper id: 25204

Abstract:

Plants play a crucial role in human life, providing food, medicine, and industrial value. Citrus maxima, a member of the Rutaceae family, are traditionally used for its medicinal properties, especially its peel, which is rich in flavonoids such as Naringin and Naringenin. This study aimed to evaluate the phytochemical profile and antioxidant potential of C. maxima peel extracts compared to Carnation and C. sinensis. Peels were collected, dried, and subjected to Soxhlet extraction using methanol. Phytochemical screening, DPPH radical scavenging assay, TLC, HPLC, and LC-MS were performed to identify and quantify flavonoids. Methanolic extracts of C. maxima showed a high presence of flavonoids, particularly Naringin and Naringenin. DPPH assay confirmed significant antioxidant activity. TLC and chromatographic techniques confirmed the presence and concentration of bioactive flavonoids. C. maxima peel is a potent source of natural antioxidants, particularly flavonoids, and holds promise for pharmaceutical and nutraceutical applications. Utilizing the peels can also reduce environmental waste.

On Bayesian Estimation for Modified-Weibull Distribution

Author Name: Harshita Kumawat, Pankaj Nagar

Paper id: 25205

Abstract:

This article studies the Bayesian estimation in newly developed Modified-Weibull (M-W) Distribution. The M-W distribution is suitable for bath-shaped and non-decreasing hazards. The bayes estimators of shape and scale parameters and reliability characteristics of the distribution are obtained under squared error loss function (SELF). Furthermore, the bayes estimates are comupted using Tierney Kadane approximation and Markov Chain Monte Carlo techniques. The High Posterior Density (HPD) Credible intervals of parameters are also obtained. A Monte Carlo simulation study is performed to compare the Maximum likelihood estimators and Bayes Estimators. Under certain circumstances, the proposed estimators are observed more efficient than already existing estimators. Finally, a real data set is used for illustrative purposes.