Kirubel Biruk (PhD)M ulu Y emiamrew2026-11-132026-11-132026-10https://etd.aau.edu.et/handle/123456789/9511The aim of this study is to evaluate determinants in terms of stakeholder collaboration, data quality, scenario planning, staff expertise, and feedback mechanisms on inventory optimization , where forecasting accuracy acts as a mediator, at the Et hiopian Disaster Risk Management Commission (EDRMC). Using an explanatory research design along with quantitative research design was adopted , and the required data was collected from 201 logistic and forecasters in eleven departments. The data collected f rom structured questionnaire were analyzed using descriptive analysis, Pearson correlation, and mediation analysis by 5,000 re sampling of the data set. All the five determinants together explained 68.7% of the variance in inventory optimization (R ² = 0.68 7, p < 0.001). The variables were ordered based on their ability to predict inventory optimization as follows; staff expertise (β = 0.281, p < 0.001), data quality (β = 0.223, p < 0.001), stakeholder collaboration (β = 0.210, p < 0.001), feedback mechanism (β = 0.203, p < 0.001), and scenario planning (β = 0.140, p = 0.002). The forecasting accuracy of the organization proved to have a partial mediating role in all the five predictor variables relationship with indirect effect (B = 0.072, SE = 0.011, p < 0. 001; bootstrapped CI [0.031, 0.118]). Organizational factors explained 40.4% of the variance in the dependent variable (R ² = 0.404, β = 0.636, p < 0.001). Descriptive results revealed that scenario planning was the least influential factor (M = 3.40) and u se of single deterministic forecast was high (M = 2.71). EDRMC embed scenario based forecasting into policy by requiring at least three alternative demand scenarios in every inventory plan. This reduces reliance on single deterministic forecasts (M = 2.71) and prepares the organization for uncertainty.en-USKeywords: Inventory OptimizationForecasting AccuracyHumanitarian Supply ChainOrganizational DeterminantsMediation AnalysisDisaster Risk ManagementEthiopiaThesis