Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study introduces a long-memory reservoir computing framework tailored for dengue forecasting, addressing the challenges posed by short and noisy incidence series. By integrating fractional-differencing dynamics and wavelet smoothing into Echo State Networks, the proposed models (fESN and wESN) effectively capture long-range temporal dependencies while maintaining the ability to model nonlinear dynamics. The results demonstrate that both fESN and wESN significantly outperform existing statistical and deep learning approaches across various datasets and forecasting horizons, while also providing calibrated uncertainty intervals through conformal prediction.
Long-memory reservoir computing models can dramatically enhance dengue forecasting accuracy, outperforming traditional methods and deep learning approaches in data-scarce environments.
Accurate dengue forecasting is crucial for public health planning, but remains challenging because incidence series are often short, noisy, non-stationary, nonlinear, and often affected by long-range temporal dependence. Fractional differencing in Autoregressive Fractionally Integrated Moving Average (ARFIMA) helps balance non-stationarity and persistence, but its linear structure limits its ability to capture nonlinear dynamics. Deep neural networks can model nonlinear patterns, but usually require large training samples and do not explicitly encode statistical long memory. Echo State Networks (ESNs), a widely used reservoir computing framework, are attractive in this setting because they retain nonlinear recurrent dynamics while training only a simple readout, making them suitable for data-scarce scenarios. However, standard ESNs lack long-term memory from a time-series perspective. This study proposes a long-memory reservoir computing framework that integrates dedicated long-memory and short-memory ESN reservoirs with a ridge-regression readout. We introduce two variants: Fractional ESN (fESN), which incorporates fractional-differencing dynamics into the reservoir to encode long-range dependence directly, and Wavelet ESN (wESN), which extracts stable low-frequency components through wavelet smoothing before modeling them with a memory-aware reservoir. We establish theoretical guarantees for closed-loop reservoir dynamics, showing that standard ESNs induce short-memory processes under mild conditions, whereas the proposed long-memory reservoirs generate polynomially decaying dependence consistent with statistical long memory. Across multiple dengue datasets and forecasting horizons, fESN and wESN outperform statistical and deep learning baselines. Combining conformal prediction with fESN and wESN provides distribution-free calibrated uncertainty intervals.