A hybrid neutrosophic under soft topological and ANN-based approach to decision-making under uncertainty

This paper introduces a new hybrid decision-making framework that uses \(\mathcal {N}_{\mathfrak {s}}^{\mathfrak {u}}\) -compact topological structures and artificial neural networks to solve industrial challenges which occur under conditions of uncertainty and limited data availability. The proposed model uses neutrosophic under soft sets to represent truth, indeterminacy, and falsity components which reflect the unpredictable and inconsistent behavior of actual chemical storage facilities. The sum-correlation function enables the conversion of neutrosophic data into numerical value assessments which organizations can use to assess and prioritize their storage facilities. The concept of \(\mathcal {N}_{\mathfrak {s}}^{\mathfrak {u}}\) -compactness helps identify the essential monitoring systems needed for complete system monitoring while ensuring efficient resource use. The proposed method includes an artificial neural network model which learns how neutrosophic inputs connect to decision score outputs to demonstrate its reliability. The results show that theoretical values and artificial neural network predictions exhibit strong alignment which indicates that the proposed framework is reliable and produces consistent results. The hybrid model functions as an intelligent decision-support system which effectively supports chemical industrial safety assessments and risk evaluations and other fields that deal with uncertain situations.