Multi-variable regression analysis for the solid waste generation in the State of Kuwait

• Published in: Process safety and environmental protection Vol. 119; pp. 172 - 180
• Main Authors: Al-Salem, S.M., Al-Nasser, A., Al-Dhafeeri, A.T.
• Format: Journal Article
• Language: English
• Published: Rugby Elsevier B.V 01.10.2018; Elsevier Science Ltd
• Subjects: Agricultural economics; Agricultural management; Agricultural production; Commercial buildings; Construction; Construction costs; Data models; Dependent variables; Developing countries; Generation; Independent variables; Kuwait; LDCs; Modelling; Population (statistical); Prediction models; Regression; Regression analysis; Regression models; Reliability analysis; Residential buildings; Solid waste; Solid wastes; Statistical analysis; Waste disposal; Waste management; Kuwait; KM; WCons; SW; tpa; SWM; Regression; tpd; WComm; HSW; MENA; MSW; SD; WH; GDP; SE; VIF; Generation; Solid waste; Ktpa; Modelling; Kuwait; WAgri; B&C
• ISSN: 0957-5820; 1744-3598
• DOI: 10.1016/j.psep.2018.07.017

[Display omitted] •In this work, multi-variable regression models were developed for various solid waste components on a state level.•The developed regression models can predict individual waste components with high accuracy.•High accuracy was also noted between predicted and input parameters for the models developed.•This work solves future problems for waste mitigation and strategy development. Accurate prediction of solid waste (SW) generation is considered an important aspect of waste management. It plays a major role in strategy development especially in developing countries. In this work, six independent variables related to the state of Kuwait were used as inputs for the development of multi-variable regression models. The aim was to predict SW generation rates from a number of sectors within the country, namely the domestic, commercial, building and construction (B&C), and agricultural ones. The variables included comprised the total population of the country, gross domestic product (GDP) index, construction area, cost of utilised constructed agricultural area and total agricultural production requirements. Statistical analysis was used to confirm the reliability of the regression models developed. The results indicated that predications were highly accurate with standard errors (SE) ranging between 3.52% and 10.46% for the indicators of the multiple regression predictive models. Multiple-variable regression models developed showed mean standard errors ranging between 0.125 and 1.09% for the dependent variables considered. The developed regression models can be used to predict individual SW components which could be used by decision makers when devising measures and policies for long-term SW management strategies.