Conference Agenda
Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
|
Daily Overview |
| Session | |
|
Poster session: Poster session with drinks and snacks Location: HSB3 | |
    | |
Poster
Urban flood prediction and mapping using Machine Learning and Deep Learning 1: Institute for Artificial Intelligence R&D of Serbia, Serbia; 2: Faculty of Civil Engineering, University of Belgrade, Serbia; 3: School of Civil and Environmental Engineering, University of New South Wales (UNSW), Sydney, Australia Urban floods pose serious risks to high-density areas and complex infrastructures, causing social, economic, and environmental damages. Traditional physics-based flood prediction methods are slow and computationally intensive, limiting real-time forecasting. Since 2017, machine learning (ML) and deep learning (DL) models have emerged to accelerate predictions (after initial lengthy training), though approaches vary widely. This study analyzes recent ML and DL efforts (e.g., Decision Trees, Support Vector Machines, Convolutional Neural Networks, Long-Short Term Memory, etc.) for predicting flash flood timing, extent, and urban impact. By reviewing literature datasets on weather, terrain, and historical flood records, we identify key inputs and metrics guiding future research (what to do and what not to do). Common input data include rainfall, slope, elevation, and proximity to rivers and roads, with images also used in DL models. Performance metrics such as precision (0.7-0.98), accuracy (0.64-0.98), and AUC (0.69-0.99) indicate strong model performance. However, most models rely on synthetic data, and challenges persist in validating results with real-world measurements. | |