Come along to this in-person seminar presented by Dr Mahdi Abolghasemi on enhancing forecasts for optimal decision making.
Forecasting plays a pivotal role in managerial decisions faced with uncertainty. Forecasts are often not the final goal but are used as an input in decision making models. This paradigm, known as “predict and optimise,” uses forecasts alongside other parameters to derive optimal decisions and have found numerous applications in real world problems, e.g., forecasting sales and determining optimal inventory, forecasting demand and scheduling set of activities. Literature suggests more accurate forecasts do not necessarily lead to better decisions but there may be a correlation between them. However, it is not trivial how one can change the forecasts to obtain better decisions. I will present our winning method from the IEEE 3rd technical challenge to predict and optimise, where the goal was to forecast the electricity demand and solar power, and accordingly solve a scheduling problem to minimise total electricity cost. I will further delve into enhancing forecasts via loss function to improve the decisions and touch upon recent advances in merging forecasting with optimisation via machine learning.
Dr. Mahdi Abolghasemi is a lecturer in Data Science at the School of Mathematics and Physics at The University of Queensland. His research interests and expertise lie in time series forecasting, predictive analytics, and machine learning, with applications in supply chain management and renewable energy optimisation. Mahdi serves as a consultant and chair for the International Institute of Forecasters and sits on the editorial board of the International Journal of Forecasting.
Details:
| Location: | GP-Z207, Gardens Point Z Block, Room 207 |
| Start Date: | 31/08/2023 [add to calendar] |
| Start Time: | 10 AM |
| End Date: | 31/08/2023 |
| End Time: | 11 AM |