Time Series Forecasting for Water Applications Using N-HiTS

Type of project

M.Sc.

Contact

Rocco Palmitessa (rpal@dhigroup.com)

Topic

N-HiTS is a recent deep learning model for time series forecasting that has shown state-of-the-art performance across many domains. This project explores how N-HiTS performs on typical water-related time series such as wastewater inflows, harbour water levels, or urban runoff. The student will compare deep learning performance with traditional time series forecasting methods across 3–5 real-world datasets.

Relevance to DHI

Accurate time series forecasting is a critical component in many DHI solutions, including early warning systems and operational planning. Leveraging modern neural forecasting techniques like N-HiTS could improve accuracy and robustness, particularly for irregular or nonlinear patterns.

What is the student's contribution

Apply N-HiTS and traditional methods to selected forecasting problems. Compare accuracy, robustness, and operational feasibility. Deliverables: code, comparative analysis, and guidelines for model selection.