Harith, S. M.2024-03-122024-03-122023Harith, S. M. (2023). Modelling socio-economic and demographic determinants of food poverty among household heads in Zanzibar, (Master's Dissertation) The University of Dodoma.https://repository.udom.ac.tz/handle/20.500.12661/4265In Zanzibar, food poverty among households has become a serious problem, especially in rural areas where a sizable population lives in food poverty. Food poverty among household members significantly contributes to nutritional quality, possibly leading to inadequate nutrition and posing dangers to physical, mental, and financial health. The Zanzibar Household Budget Survey of 2019-2020, which comprised 2788 household heads, was used in this study. This study's objective was to investigate the socio-economic and demographic variables that affect household heads' access to food in Zanzibar. Kilocalories from food consumption per equivalent adult were used as a proxy measure with a baseline value of 2200kc (equal to 1571.109088396 Tanzanian shillings) to evaluate the food poverty status among adults in Zanzibar. The socio-economic and demographic correlates of food poverty in Zanzibar families were identified using a designed based logistic regression model. According to the survey analysis, just 6.00% of household heads were assessed to be food insecure, while 94.00% were food secured. Household size, the level of education of the household heads, total monthly spending, the category of unemployed people, and location (urban or rural) were all significant predictors of food insecurity among household heads in Zanzibar using designed based binary logistic regression model. However, according to the multilayer perceptron neural network model, household size was the most significant predictor of food poverty. Other significant predictors included total monthly domestic spending, the employment status of household heads, and marital status. Both the multilayer perceptron neural network and the designed-based binary logistic regression model performed well. Nevertheless, the designed-based binary logistic regression model was outperformed by the multilayer perceptron neural network. The income levels and economic prospects of Zanzibar households should be improved through action. This can be done by implementing strategies that improve income and lower food insecurity, such as job development, business support, sustainable livelihoods, and vocational training. The improvement of family food security depends heavily on governmental and non-governmental organizations like the Food and Agriculture Organization (FAO) and the World Food Programme (WFP).enThesis