SIWW SPOTLIGHT 2025-IAHR2025 PROSPECTUS
[DEADLINE: 15 JAN 2025]
SUBMIT YOUR ABSTRACTS FOR SPECIAL SESSIONS NOW

Workshop: Machine Learning Approaches for Hydrologic Modelling and Data Quality Assessment

Convenor: Prof. Ramesh S. V. Teegavarapu, USA

Objective(s)
This workshop aims to introduce the concepts of Machine Learning (ML) approaches for hydrologic modelling and data quality assessment and improvement. The workshop will focus on the fundamentals of ML techniques for hydrologic forecasting, data quality improvement, and approaches supporting water resources management.

Workshop Topics
The workshop will cover and discuss the following topics:

  • Data Analysis Methods: Exploratory and Confirmatory, Big Data Essentials, Hydroinformatics and Hydroanalytics.
  • Introduction to Statistical and Machine Learning (ML): Fundamentals
  • Overview of ML techniques: Classification, Clustering, predictive modelling, pattern recognition, anomaly detection, dimensionality reduction.
  • Tree-based Approaches: Decision trees, Regression, Model Trees, and Related Ensemble Approaches.
  • Clustering (K-means and variants), Classification (KNNs and variants), Support Vector Machines
  • Regression Approaches (K-NN, Gaussian Kernel Regression, Logistic Regression)
  • Artificial Neural Networks (ANNs) and Deep Learning (DL)
  • Application of ML techniques for Hydrologic Modelling, Data quality assessment, missing data estimation, and prediction.
  • Evaluation of ML-based Models: Calibration and Validation
  • Emerging ML Approaches and Directions for Future Research

The workshop would cover the in-depth discussion of concepts and review applications of the ML techniques based on several case studies.

Expected Benefits
The participants are expected to learn more about the ML tools and explore their functioning. In addition, they would be introduced to generic techniques for model calibration, validation, predictor selection, and model evaluation. At the end of the workshop, the participants are expected to acquire sufficient knowledge to appreciate the different ML techniques and be able to select the best techniques to solve real-world hydrologic problems.