Uncertainty Analysis in Prediction Intervals Using Neutrosophic Numbers

Abstract

This study analyzes uncertainty in prediction intervals by applying neutrosophic numbers to quality of life forecasts based on characteristics such as compensation, job hours, years of experience, and academic level. Three examples were studied, with indeterminacy values ranging from 0.589 to 0.628. These statistics show that between 58.9% and 62.8% of the prediction intervals are affected by uncertainty. Higher indeterminacy levels reflect a stronger effect of external influences or unpredictability within the data. The analysis demonstrates that as variables like compensation and labor hours grow, so does the level of uncertainty in the projections. Neutrosophic numbers provide a valuable framework for quantifying this uncertainty, providing for a better understanding of the model's limitations and the unpredictable nature of the dataset.