International Journal of Information Technology and Applications, Vol. 2, No. 1, pp. 1-14, March 2025.
Abstract: To enhance the prediction accuracy rate of the fuzzy inference system (FIS) in nonlinear time series, the Long Short-Term Memory (LSTM) network is integrated into the FIS to improve the ability of neural networks to process the fuzzy information. On this basis, the Quantum Echo State Network (QESN) is added and optimized by the particle algorithm to predict the nonlinear time series. The results indicate that the proposed new self-evolving interval type II LSTM (eit2FNN-LSTM) can realize single-point prediction; also, it can be used for granular-level prediction by combining with information particles, thereby achieving long-term prediction; besides, the QESN can predict the time series well, and its accurateness rate is higher than the benchmark algorithm. Therefore, the FIS based on quantum computing and the new eit2FNN-LSTM can effectively improve the prediction accuracy rate in the nonlinear time series, providing new ideas for prediction problems in other fields.