Your Cart

Your cart is empty

Home  >  Volume 54

12. NON LINEAR TIME SERIES MODELLING AND CHANGE POINT DETECTION OF MUNICIPAL SOLID WASTE GENERATED IN LAGOS STATE, NIGERIA by J. F. Ojo and Olaoyenikan O. B. Volume 54 (January 2020 Issue)
Sale price: $0.00
Quantity:

NON LINEAR TIME SERIES MODELLING AND CHANGE POINT DETECTION OF MUNICIPAL SOLID WASTE GENERATED IN LAGOS STATE, NIGERIA

J. F. Ojo and Olaoyenikan O. B.


Department of Statistics, University of Ibadan, Ibadan, Nigeria.


Abstract


In this study, we analyse municipal solid waste generated in Lagos State using non-linear time series model and change point detection technique. This is necessary because so much attention have been given to linear models when it comes to analysing municipal solid waste generated in Lagos State. Change point detection technique was used to measure the intervention of government on volume of municipal solid waste generated in Lagos State. We fitted non-linear time series model particularly bilinear time series model and compared it with linear model namely autoregressive integrated moving average model. The data used for this study exhibited non-stationary having conducted unit root test. Using Akaike information criterion the appropriate model fitted for nonlinear and linear models respectively were of the order (6, 1, 0) and (6, 1, 0, 1, 1).There was a single change point in the volume of solid waste generated between 1980 and 2014 and this occurred in 1987 with an upward shift. This change could be caused by increase in population and efficient service of the waste authorities. The non-linear model performed better than the linear model with a lower residual variance and root mean square error. We recommend non-linear times series model particularly bilinear model for predicting municipal solid waste that will be generated in Lagos State.


Keywords: Solid waste, Change point, Pettitt approach, Intervention, Time series models

Close
Loading...