General Notices

The Department of Mathematics and Statistics invites the campus community to attend a Ph.D. Seminar presented by Ms. Aruna Rajballie of the Faculty of Science and Technology.

Title: Hybrid Time-Series Models for Forecasting Monthly Water Consumption Across Multiple Demand Sectors

This seminar examines advanced statistical and hybrid modelling approaches for forecasting water consumption in Trinidad, with implications for effective water-demand management and planning.

Seminar Details

-          Presenter: Ms. Aruna Rajballie

-          Department/Faculty: Department of Mathematics & Statistics, Faculty of Science & Technology

-          Date: Friday, 23 January 2026

-          Time: 10:00 a.m. – 11:00 a.m.

-          Venue: Department of Mathematics and Statistics Seminar Room, 2nd Floor, Natural Sciences Building, FST

-          Online (Zoom): https://bit.ly/3N9ZUXU

-          Meeting ID: 930 1270 7127                                

-          Passcode: 294129

Abstract

Urbanisation in Trinidad has intensified over recent decades, contributing to a growing national demand for water. Accurate forecasting of water consumption is essential for understanding usage patterns across economic sectors and for supporting effective water-demand management. This study evaluates the performance of seasonal ARIMA (SARIMA), exponential state-space (ETS), artificial neural network (ANN) models, and their hybrid combinations in forecasting monthly water consumption across major demand categories in Trinidad. Model performance is assessed using RMSE, MAE, and MAPE. Results indicate that hybrid modelling approaches provide superior accuracy for most sectors, highlighting their value in improving national water-demand forecasting.

All interested persons are welcome to attend, either in person or online.