Nickel Price Forecast Based on the LSTM Neural Network Optimized by the Improved PSO Algorithm

Shao, Bilin; Li, Maolin; Zhao, Yu; Bian, Genqing

doi:10.1155/2019/1934796

Cited by 39 publications

(17 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to the performance metrics MAPE, RMSE, MAE, and R 2 , the proposed model still achieves the best performance among the four models. Table 3 shows that Wang et al [28] and Shao et al [29] could improve the forecasting results of LSTM. However, the proposed IPSO-LSTM model still has the best performance over all four models.…”

Section: Table 2 Comparison Of the Forecasting Results Corresponding To The Different Evaluation Metricsmentioning

confidence: 98%

“…7 that the IPSO-LSTM forecasting is very close to the true curve. To further verify the forecasting capability of the IPSO-LSTM model, we compare it with the PSO-LSTM and the other improved variable inertia weight methods of Wang et al [28] and Shao et al [29]. Except for the look-back period, the other parameter settings for those models are inherited from the original papers.…”

Section: Table 2 Comparison Of the Forecasting Results Corresponding To The Different Evaluation Metricsmentioning

confidence: 99%

“…As the parameters of LSTM are usually set heuristically, they are unlikely to be optimal. Recent advances indicate that the parameters of LSTM can be updated dynamically using PSO algorithm [28], [29]. However, PSO algorithm often falls into local optimum during iteration.…”

Section: A Ipso-lstm Modelmentioning

confidence: 99%

See 2 more Smart Citations

A Novel Improved Particle Swarm Optimization With Long-Short Term Memory Hybrid Model for Stock Indices Forecast

Liew

Yang

2021

IEEE Access

View full text Add to dashboard Cite

Stock market volatility has a significant impact on many economic and financial activities in the world. Forecasting stock price movement plays an important role in setting an investment strategy or determining the right timing for trading. However, stock price movements are noisy, nonlinear, and chaotic. It is difficult to forecast stock trends for improving return on investment. Here, we proposed a novel improved particle swarm optimization (IPSO) and long-short term memory (LSTM) hybrid model for stock price forecasting. An adaptive mutation factor was used as a parameter for model optimization to avoid premature convergence to a local optimum. Furthermore, we presented a nonlinear approach to improve the inertia weight of the particle swarm optimization and then used the IPSO model to optimize the hyperparameters of LSTM. The experimental results showed that the proposed model outperformed the other related baseline models: support-vector regression, LSTM and PSO-LSTM on the Australian stock market index. These results indicated the proposed model possesses high reliability and good forecasting capability.INDEX TERMS long-short term memory network, time series forecasting, stock indices forecasting, hybrid feature selection, deep learning

show abstract

Section: Table 2 Comparison Of the Forecasting Results Corresponding To The Different Evaluation Metricsmentioning

confidence: 98%

Section: Table 2 Comparison Of the Forecasting Results Corresponding To The Different Evaluation Metricsmentioning

confidence: 99%

See 1 more Smart Citation

A Novel Improved Particle Swarm Optimization With Long-Short Term Memory Hybrid Model for Stock Indices Forecast

Liew

Yang

2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The improved PSO-LSTM model is compared with the conventional LSTM and the integrated moving average autoregressive model (ARIMA). The results show that compared with the standard PSO, the improved PSO has a faster convergence rate and can improve the prediction accuracy of the LSTM model effectively [17].…”

Section: Lstm Forecasting With Pso Approachmentioning

confidence: 97%

Timeseries Forecasting using Long Short-Term Memory Optimized by Multi Heuristics Algorithm

Hendri,

Wibowo*,

Sari

2019

IJRTE

View full text Add to dashboard Cite

Forecasting future price of financial instruments (such as equity, bonds and mutual funds) has become an ongoing effort of financial and capital market industry members. The most current technology is usually applied by high economic scale companies to solve the ambitious and complicated problem. This paper presents optimization solution for a deep learning model in forecasting selected Indonesian mutual funds' Net Asset Value (NAV). There is a well-known issue in determining a deep learning parameters in LSTM network like window timestep and number of neurons to be used in getting the optimal learning from the historical data. This research tries to provide solution by utilizing multi-heuristics optimization approach consists of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to determine the best LSTM's network parameters, namely window timesteps and number of neurons. The result shows that from the nine selected mutual funds, PSO outperforms GA in optimizing the LSTM model by giving a lower Root Square Mean Error (RMSE) by 460.84% compared to GA's. However, PSO took a longer execution time by 1.78 times of GA's. This paper also confirms that based on RMSE for both training and evaluation dataset, equity mutual fund's forecasted NAV has the highest RMSE followed by fixed income mutual fund's forecasted NAV and money market mutual fund forecasted NAV.

show abstract

“…At the same time, weight and bias values are optimized by particle swarm optimization (PSO) algorithm for Nickel price forecasting. 32 The PSO-LSTM and the ARIMA model can improve MAPE by 9% and 13% compared to conventional LSTM, respectively. Also, residual neural network (ResNet) has widely used in computer vision and pattern recognition.…”

Section: Introductionmentioning

confidence: 96%

An electricity price interval forecasting by using residual neural network

Chaweewat

Singh

2020

Int Trans Electr Energ Syst

View full text Add to dashboard Cite

Summary This article proposed a new electricity price interval forecasting method based on a novel Residual Neural Network (ResNet) for the electricity price interval forecasting. The significant outcome of the ResNet model was that the model performs excellently on normal and spike price interval forecasting in accuracy and reliability point of view. The proposed ResNet was consisting of two network layers. The first neural network layers were probabilistic normal, high, and spike prices prediction part. The Lower and Upper Bound Estimation (LUBE) formulates the price interval forecasting from the output of the second neural network layers. The LUBE methods included Quantile Regression and Mean and Variance estimation. The proposed forecasting models were demonstrated with the GEFCom2014 dataset. The dataset is consisting of 15 tasks for electricity prices forecasting. The results of proposed ResNet models compared with GEFCom2014's benchmarks, Quantile Regression Average, and Multilayer Perceptron Network approaches. The performances of forecasting models are evaluated in terms of accuracy and reliability metrics by Pinball Loss Function and Coverage Width‐based Criterion (CWC), respectively. Also, this article shows the effect of an increase in confidence level, which could generate lower CWC values and represent high reliability's satisfaction.

show abstract

Nickel Price Forecast Based on the LSTM Neural Network Optimized by the Improved PSO Algorithm

Cited by 39 publications

References 16 publications

A Novel Improved Particle Swarm Optimization With Long-Short Term Memory Hybrid Model for Stock Indices Forecast

A Novel Improved Particle Swarm Optimization With Long-Short Term Memory Hybrid Model for Stock Indices Forecast

Timeseries Forecasting using Long Short-Term Memory Optimized by Multi Heuristics Algorithm

An electricity price interval forecasting by using residual neural network

Contact Info

Product

Resources

About