Skip to main content

Introduction

Author: Cansu Moran

Overview​

Reservoir computing is a powerful and efficient machine learning technique designed for time-series prediction tasks. It is particularly well-suited for forecasting and dynamic system analysis. In this project, reservoir computing is applied to predict the next 90 days of FCS (Food Consumption Score) and RCSI (Reduced Coping Strategies Index) values for various countries based on their historical data.

Purpose​

The goal of this project is to enhance decision-making and resource allocation by providing accurate, up-to-date forecasts of FCS and RCSI indices. Predictions are updated daily to account for new data and ensure the latest information is always available.

Key Features​

  • Daily predictions using updated datasets.
  • 90-day forecasts for FCS and RCSI indices, including their upper and lower bounds.
  • Integration with a database for storing and updating predictions.
  • Accessible predictions through a dedicated API endpoint for frontend usage.