AGRICULTURAL DATA SOURCE SELECTION FOR CROP YIELD ANALYSIS USING HESITANT FUZZY ROUGH MULTI-ATTRIBUTE DECISION-MAKING METHOD

Agricultural productivity largely depends on the quality and reliability of data used for crop yield analysis.Multiple data sources, such as soil databases, satellite imagery, and government agricultural statistics, are available, but selecting the most suitable source remains a challenging task due to uncertainty, incompleteness, and variability in data quality.Traditional data selection approaches often rely on fixed rules or simple weighting mechanisms, which may not effectively handle ambiguity in expert judgements.To address this issue, this study proposes a Hesitant Fuzzy Rough Multi-Attribute Decision-Making (HFR-MADM) framework for optimal agricultural data source selection.The proposed method integrates hesitant fuzzy sets to model uncertainty in expert evaluations and rough set theory to manage vagueness in decision boundaries.Multiple evaluation criteria, such as accuracy, reliability, timeliness, accessibility, and cost, are considered for ranking alternative data sources.Experimental analysis demonstrates that the proposed approach improves decision consistency and provides a more robust ranking compared to conventional selection techniques.The developed framework supports informed decision-making and enhances the reliability of crop yield analysis systems.