The Tokyo University of Science in Japan shares that a recent “breakthrough” will improve AI applications in finance, weather forecasting, and speech recognition.
The innovation involves reservoir computing (RC). Researchers Dr. Masanobu Inubushi and Ms. Akane Ohkubo at the university have developed a “novel readout method for reservoir computing” that is described as significantly improving prediction accuracy. By applying the mathematical theory of synchronization, deeper patterns in time-based data are said to emerge.
Reservoir computing is said to be improved by reducing the amount of computation needed; it is faster and more computationally efficient. The innovation is said to be inspired by how the brain works. It is said to be especially good at predicting complex systems.
Prediction applications are key tools in financial services. The researchers tests are said to generate “notable improvements in accuracy, along with an unexpected enhancement in robustness, both in short-term and long-term predictions, compared to conventional RC.”
If you are interested, you may read the full report here.
