向量化複數模糊神經方法在時間序列智慧型預測之研究──使用PSO、WOA與RLSE

A Study on Time Series Intelligent Forecasting by a Vectorized Complex Neuro-Fuzzy Approach Using PSO, WOA and RLSE

本研究針對時間序列問題，建構一個向量化複數模糊神經系統。模型引入自我建構概念，以減法分群法利用輸入資料提供的資訊決定初始參數。使用複數模糊集合，使模型可進行多目標預測。模型參數學習提出PSO-WOA-RLSE（Particle Swarm Optimization-Whale Optimization Algorithm-Recursive Least Squares Estimator）混合式演算法，引入分治法概念，將參數分為兩個集合，分別為模糊系統前鑑部及後鑑部參數；前鑑部參數由PSO-WOA調整；後鑑部參數由RLSE調整。測試階段時，預測完資料可視為已知資料，利用RLSE的特性，在預測完一對資料後，用極低成本更新參數，可降低產生過度擬合機率。為驗證所提出模型之預測效能，本研究使用三個實驗進行試驗。實驗結果表明本研究所提出模型與其它研究文獻相比擁有良好的預測效能。

For financial time series problems, this study proposes a vectorized complex neuro-fuzzy system (VCNFS). The proposed model is based on a neuro-fuzzy framework, where there are several If-Then rules constructed with complex fuzzy sets (CFSs) in terms of a neural network. The premise parts of If-Then rules are determined initially by the subtractive clustering (SC) algorithm with the information given by data, while the consequent parts are polynomial functions. Thus, the model is basically realized by the self-organization and data-driven concept. The use of CFSs enables the proposed model to perform multi-target prediction. For model parameter learning, a novel hybrid learning algorithm is proposed, called the PSO-WOA-RLSE, which combines the particle swarm optimization (PSO) with the whale optimization algorithm (WOA) and the recursive least squares estimator (RLSE). The PSO-WOA-RLSE algorithm uses the divide-and-conquer principle, where the PSO and WOA cooperate with each other to adjust the parameters of the premise parts of If-Then rules, and the RLSE to adjust the parameters of the consequent parts, so that the proposed model can converge quickly with good accuracy. Parameter learning by the PSO-WOA-RLSE algorithm separates the parameters that need to be optimized in a single algorithm, so to reduce the burden of the algorithm, and improves the performance of the model, in terms of quick convergence and optimization accuracy. In the testing phase, the use of RLSE to update parameters with known data after prediction can reduce the probability of overfitting and thus can upgrade the prediction performance of the system. Three experiments are designed in this study, all of which use financial timeseries datasets to verify the performance by the proposed multi-target forecasting model. The experimental results show that the model proposed in this study has good prediction performance, compared with other research literatures.