Date of the expedition
From 01/07/2023 to 31/12/2023
Deep Learning Based Harmful Algal Blooms Prediction
University of Iowa, Hydroinformatics Lab
Dr. Mermer is now working as a Senior Research Fellow at University of Iowa HydroInformatic Lab (UIHILab). He received his PhD. in Physics/Applied Engineering from University of Iowa. Dr. Mermer’s research focuses on data driven approach, AI model development, feature engineering and sensor system. Currently, he is conducting studies on ML/DL model development for prediction/anomaly detection and integration of explainable AI. He is also working on digital twin system and next-gen web technologies.
Harmful algal blooms (HABs) are a serious environmental problem for all nations. HABs have several detrimental consequences for the environment, the economy, water quality, wildlife, fisheries, public health, and recreational activities. While various HAB causes are known, the precise development of toxin-producing algae and the factors contributing to their mass remain uncertain. Although climate change is known to have an influence on HABs, the link between atmospheric factors and harmful algal growth is still unclear. Current studies lack data and understanding of the main causes.
To address this gap, this project aims to develop a prediction model using deep learning, identify driving factors through explainable AI algorithms and provide a benchmark dataset that follows the FAIR (findability, accessibility, interoperability, and reuse) data principles.
In the first three months of the projects, following results are achieved and we are on track to meet our project objectives.
- Extensive literature review
- Data collection and preparation comprehensive dataset
- Survey of DL models and algorithms
- Survey of Explainable AI (XAI) model for interpreting the model output
- Planning of journal articles and Conference presentations
Impact of the Fellowship
In the first three months of the projects, the NGI Enricher fellowship has had a profound impact on the project itself and our professional growth.
- Development/advancement of innovative technologies: exposure to cutting-edge techniques such as DL, XAL, ML
- Testing technologies (demo, pilot…): applying our models on Lake Erie as pilot ecosystem.
- Sound Scientific Validation: DL algorithms, data regression, XAL methods
- Strengthening research collaboration with the US/Canada: Working with different research lab, research institute and research community.
- Building solid connections and partnerships in Europe and in the US/Canada: Working with nonprofit organization (MBDH), research community (CUAHSI)
- Fundraising (proposals to public organisations) – indicate, in the US/Canada or Europe: preparing new projects for Climate Change AI and MBDH
- Paper submission for further publication – indicate, only EU author(s), or jointly with the host organisation: preparing a journal article with the host organisation
- Conference attendance with paper/poster/ proceedings: preparing a journal article with the host organisation