Researchers have developed a novel long-memory reservoir computing framework to improve dengue forecasting, particularly in data-scarce scenarios. The proposed system integrates dedicated long-memory and short-memory Echo State Networks (ESNs) with a ridge-regression readout. Two variants, Fractional ESN (fESN) and Wavelet ESN (wESN), were introduced, with fESN incorporating fractional-differencing dynamics and wESN using wavelet smoothing to capture long-range dependencies. Both variants demonstrated superior performance compared to statistical and deep learning baselines across various dengue datasets and forecasting horizons. AI
IMPACT This research could lead to more accurate public health planning by improving disease forecasting models.
RANK_REASON The cluster contains a research paper detailing a new methodology for forecasting, published on arXiv.
- autoregressive fractionally integrated moving average
- Dengue virus
- Echo State Networks
- Fellow of the Entomological Society of New Zealand
- Fractional ESN
- Long-Memory Reservoir Computing
- Tanujit Chakraborty
- Wavelet ESN
- Wesner
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
