A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Gold Price Prediction Using Skewness and Kurtosis Based Generalized Auto-regressive Conditional Heteroskedasticity Approach with Long Short Term Memory Network
Mining companies, investors, financial institutions and related firms require a precise prediction approach for investigating the fluctuations in the gold prices, and taking the correct decisions. But, the future gold price data belongs to the time series forecast, and it is complex to predict the price because of noisy, chaotic and non-stationary features of data. The existing researches failed to consider the volatility information and impact of data in the upcoming value while performing the prediction. The important goal of the research is to predict the price of gold by developing an effective classifier. In this research, the Skewness and Kurtosis Based Generalized Auto-Regressive Conditional Heteroskedasticity approach (SKGARCH) is proposed for enhancing the prediction of gold price based on the deep learning classifier with volatility information. The deep learning classifier namely long short term memory (LSTM) network is considered for forecasting the gold price. The skewness and kurtosis parameter of SKGARCH are used for analyzing the distorted measures in the data distribution and for examining the tailedness probability i.e., continuation of previous data’s impact in the upcoming value. Moreover, the checking of missing values and duplicates, and min–max scaling are carried out to convert the input data in a unified manner. Therefore, the combination of SKGARCH and LSTM provides an effective prediction of gold price. The gold price prediction dataset is considered for evaluating the SKGARCH-LSTM method. The proposed SKGARCH-LSTM method is analyzed by using accuracy (ACC), mean absolute error (MAE) and root mean square error (RMSE). The existing approaches namely, convolutional neural network (CNN)-LSTM, variational mode decomposition-bidirectional gated recurrent unit (VMD-BiGRU), LSTM-Plus (LSTM-P) and eXtreme Gradient Boosting (XGBoost), are used to evaluate the SKGARCH-LSTM method. The MAE of SKGARCH-LSTM is 0.0032, which is less when compared to the CNN-LSTM and VMD-BiGRU.
Gold Price Prediction Using Skewness and Kurtosis Based Generalized Auto-regressive Conditional Heteroskedasticity Approach with Long Short Term Memory Network
Mining companies, investors, financial institutions and related firms require a precise prediction approach for investigating the fluctuations in the gold prices, and taking the correct decisions. But, the future gold price data belongs to the time series forecast, and it is complex to predict the price because of noisy, chaotic and non-stationary features of data. The existing researches failed to consider the volatility information and impact of data in the upcoming value while performing the prediction. The important goal of the research is to predict the price of gold by developing an effective classifier. In this research, the Skewness and Kurtosis Based Generalized Auto-Regressive Conditional Heteroskedasticity approach (SKGARCH) is proposed for enhancing the prediction of gold price based on the deep learning classifier with volatility information. The deep learning classifier namely long short term memory (LSTM) network is considered for forecasting the gold price. The skewness and kurtosis parameter of SKGARCH are used for analyzing the distorted measures in the data distribution and for examining the tailedness probability i.e., continuation of previous data’s impact in the upcoming value. Moreover, the checking of missing values and duplicates, and min–max scaling are carried out to convert the input data in a unified manner. Therefore, the combination of SKGARCH and LSTM provides an effective prediction of gold price. The gold price prediction dataset is considered for evaluating the SKGARCH-LSTM method. The proposed SKGARCH-LSTM method is analyzed by using accuracy (ACC), mean absolute error (MAE) and root mean square error (RMSE). The existing approaches namely, convolutional neural network (CNN)-LSTM, variational mode decomposition-bidirectional gated recurrent unit (VMD-BiGRU), LSTM-Plus (LSTM-P) and eXtreme Gradient Boosting (XGBoost), are used to evaluate the SKGARCH-LSTM method. The MAE of SKGARCH-LSTM is 0.0032, which is less when compared to the CNN-LSTM and VMD-BiGRU.
Gold Price Prediction Using Skewness and Kurtosis Based Generalized Auto-regressive Conditional Heteroskedasticity Approach with Long Short Term Memory Network
J. Inst. Eng. India Ser. B
Nallamothu, Srilekha (author) / Rajyalakshmi, K. (author) / Arumugam, P. (author)
Journal of The Institution of Engineers (India): Series B ; 105 ; 1715-1727
2024-12-01
13 pages
Article (Journal)
Electronic Resource
English
Crack diagnostics in beams using wavelets, kurtosis and skewness
| British Library Online Contents | 2014
Short-term traffic-flow forecasting with auto-regressive moving average models
| Online Contents | 2014
Auto-Regressive Neural Networks: A New Approach to Time Series Modelling
| British Library Conference Proceedings | 1998