Building a Deep Learning-Based Green Algae Prediction Pipeline for MLOps

• Research Objectives

  • The need to evolve the existing developed green algae prediction algorithm into a continuous analysis algorithm to enhance the policy utility of research outcomes
  • Development-centered research for the periodic scheduling of prediction algorithms

• Research Methods

  • Based on the chlorophyll-a prediction algorithm developed in 2019, the prediction model was reorganized to enable periodized predictions.
  • Periodization of predictions by building a data pipeline for the data collection-analysis-result derivation process of deep learning-based green algae prediction algorithm

• Pipeline Development Results

  • Building a data automation pipeline through Open API
  • Designing REST API for green algae prediction
  • Real-time chlorophyll forecast from 16 stations in 7 days
  • Providing specific examples of how artificial intelligence can contribute to solving environmental problems