| Proj No. | A3022-251 |
| Title | Multi-Objective Time Series Prediction Based on Prior Knowledge |
| Summary | Time series models often exhibit multiple periodic features, but automatically learning these periodicities can be time-consuming and may lead to incorrect feature extraction. To address this, we propose incorporating human prior knowledge into the model while adding one or more dedicated tasks for capturing periodicity. This approach will guide the model toward learning meaningful periodic patterns efficiently. For example, by introducing a monthly prediction task, the model can better capture variations across different months of the year, leading to faster convergence and improved predictive performance. This method enhances the interpretability and reliability of time series forecasting by aligning learning objectives with known periodic structures. |
| Supervisor | A/P Chau Yuen (Loc:S1 > S1 B1A > S1 B1A 12, Ext: +65 67905420) |
| Co-Supervisor | - |
| RI Co-Supervisor | - |
| Lab | Centre for Information Sciences & System (CISS) (Loc: S2-B4b-05) |
| Single/Group: | Single |
| Area: | Digital Media Processing and Computer Engineering |
| ISP/RI/SMP/SCP?: |