MAKE MOST OF THE KNOWLEDGE NETWORK, JOIN ACADEMIC RESEARCH FOUNDATION

This paper presents a novel solution methodology for a dynamic multiobjective multi-product solid transportation problem (MOMPSTP) under dual-layered uncertainty modeled via rough fuzzy variables. Traditional approaches for handling such uncertainty primarily rely on deterministic conversion using chance-constrained programming. In contrast, we develop a hybrid Particle Swarm Optimization and Genetic Algorithm (PSO-GA) to address the problem in its original uncertain form without oversimplification. Furthermore, the model incorporates dynamic credibility and trust levels, allowing real-time adjustment to changing market conditions. A comparative performance evaluation between the proposed hybrid metaheuristic and classical methods such as Weighted Sum and Ideal Point techniques is presented. The results demonstrate the superiority of the hybrid approach in obtaining a diverse and high-quality Pareto front. Sensitivity and robustness analyses are conducted to validate the adaptability of the model under fluctuating parameters. The proposed framework is particularly suitable for real-time logistics and multimodal supply chain networks with uncertain and imprecise data.

KEYWORDS: Multiobjective Optimization; Solid Transportation Problem (STP); Rough Fuzzy Variables (RFV); Particle Swarm Optimization (PSO); Genetic Algorithm (GA).

The present research examines house prices of Boston Housing dataset by using Machine Learning hybrid and ensemble methods. We compare traditional models like Random Forest with advanced techniques, including XGBoost and a hybrid ensemble of XGBoost, Random Forest, and LightGBM, which significantly improved prediction accuracy. Furthermore, the Harris Hawks Optimization (HHO) algorithm was used with LightGBM to optimize the parameters and H2O AutoML was applied to select automated models. The experimental findings indicate that these strategies surpass the existing models demonstrating their effectiveness in analyzing real estate analytics.

KEYWORDS: House price prediction; machine learning; LightGBM; HHO; Ensemble model; AutoML.

Researchers often seek to improve estimator efficiency beyond the standard mean per unit approach, which considers only the study variable’s observations. One common strategy involves incorporating data from a related auxiliary variable when drawing samples from a finite population to obtain a more accurate estimate of the study variable’s population mean. This research study introduce an innovative approach using a funnel and filter paper to isolate bias precipitates that arise in the ratio and product methods of population mean estimation under two-phase sampling. Following filtration, we examine the resulting biases and mean squared error up to the first order of approximation. Furthermore, a simulation study is performed to illustrate and validate the effectiveness of our approach.

KEYWORDS:Filtration, Biases, Two-Phase Sampling, Linear Constraints. 

The existing standard methodologies in Hydro-meteorological hazards events modeling pertaining to early warning systems, prediction, simulation, susceptibility mapping, mitigation, and assessment are crucial for understanding climate dynamics, forecasting potential climate-related hazards, and developing strategies to reduce risks and enhance resilience. These methodologies integrate various data sources, including atmospheric, oceanic, and terrestrial observations, to simulate future climate scenarios and inform policymaking and disaster preparedness efforts. The emerging Geospatial Artificial Intelligence (GeoAI) methodologies are established techniques that have continually excelled in addressing some of the above scenarios efficiently. Nevertheless, as GeoAI is in the nascent stage of study, it presents unresolved in-quiries concerning model interpretability, explainability, model generalization, and the en-hancement of its longevity (predictability). This paper thoroughly reviews the application of deep learning techniques in modeling many facets of hydro-meteorological hazards supported by a case study.

KEYWORDS: GeoAI, GeoDL, Hydro-Meteorological, Deep Learning, HMH-GeoAI, Shelf life

This article presents a method that deals with bi-index transshipment problem. A ranking method based on the point of intersection of diagonals of a generalized trapezoidal fuzzy number is applied to the fuzzy elements of a transshipment problem. An improvisation to a bi-index fuzzy transshipment problem has been done via a differently defined ranking method.

KEYWORDS: Transshipment Problem; Bi-Index; Optimization; Generalized Trapezoidal Fuzzy Numbers; Ranking.