Discovery and Prediction of Stock Index Pattern via Three-Stage Architecture of TICC, TPA-LSTM and Multivariate LSTM-FCNs

Ouyang, Hao; Wei, Xiaolu; Wu, Qiufeng

doi:10.1109/access.2020.3005994

Cited by 18 publications

(7 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, machine learning methods such as Support Vector Machine (SVM) and Gradient Boosted Decision Trees (GBDT) have become popular in stock index prediction due to their accuracy as well as less restrictive underlying assumptions [ 20 , 22 , 23 ]. Within the machine learning domain, deep learning techniques have recently drawn the attention of researchers, principally due to the high accuracy that deep learning models can reach.…”

Section: Literature Reviewmentioning

confidence: 99%

Stock index trend prediction based on TabNet feature selection and long short-term memory

2022

Self Cite

View full text Add to dashboard Cite

In this study, we propose a predictive model TabLSTM that combines machine learning methods such as TabNet and Long Short-Term Memory Neural Network (LSTM) with a complete factor library for stock index trend prediction. Our motivation is based on the notion that there are numerous interrelated factors in the stock market, and the factors that affect each stock are different. Therefore, a complete factor library and an efficient feature selection technique are necessary to predict stock index. In this paper, we first build a factor database that includes macro, micro and technical indicators. Successively, we calculate the factor importance through TabNet and rank them. Based on a prespecified threshold, the optimal factors set will include only the highest-ranked factors. Finally, using the optimal factors set as input information, LSTM is employed to predict the future trend of 4 stock indices. Empirical validation of the model shows that the combination of TabNet for factors selection and LSTM outperforms existing methods. Moreover, constructing a factor database is necessary for stock index prediction. The application of our method does not only show the feasibility to predict stock indices across different financial markets, yet it also provides an complete factor database and a comprehensive architecture for stock index trend prediction, which may provide some references for stock forecasting and quantitative investments.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Stock index trend prediction based on TabNet feature selection and long short-term memory

2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…On the one hand, the statistical vector auto-regressive algorithm (VAR) [ 5 ] and the Gaussian processes algorithm (GP) [ 6 ] are unable to explore non-linear relationships between variables. On the other hand, the LSTNet [ 7 , 8 ] and TPA-LSTM [ 9 ] methods, despite the fact that they can mine non-linear relationships, cannot explicitly determine the dependencies between any two variables. In the problem of time-series prediction, the method based on the graph neural network (GNN) [ 10 ] relies on a predefined graph structure and can obtain the relationships between variables.…”

Section: Introductionmentioning

confidence: 99%

Swing Trend Prediction of Main Guide Bearing in Hydropower Units Based on MFS-DCGNN

Li,

Xu,

Guo

2024

Sensors

View full text Add to dashboard Cite

Hydropower units are the core equipment of hydropower stations, and research on the fault prediction and health management of these units can help improve their safety, stability, and the level of reliable operation and can effectively reduce costs. Therefore, it is necessary to predict the swing trend of these units. Firstly, this study considers the influence of various factors, such as electrical, mechanical, and hydraulic swing factors, on the swing signal of the main guide bearing y-axis. Before swing trend prediction, the multi-index feature selection algorithm is used to obtain suitable state variables, and the low-dimensional effective feature subset is obtained using the Pearson correlation coefficient and distance correlation coefficient algorithms. Secondly, the dilated convolution graph neural network (DCGNN) algorithm, with a dilated convolution graph, is used to predict the swing trend of the main guide bearing. Existing GNN methods rely heavily on predefined graph structures for prediction. The DCGNN algorithm can solve the problem of spatial dependence between variables without defining the graph structure and provides the adjacency matrix of the graph learning layer simulation, avoiding the over-smoothing problem often seen in graph convolutional networks; furthermore, it effectively improves the prediction accuracy. The experimental results showed that, compared with the RNN-GRU, LSTNet, and TAP-LSTM algorithms, the MAEs of the DCGNN algorithm decreased by 6.05%, 6.32%, and 3.04%; the RMSEs decreased by 9.21%, 9.01%, and 2.83%; and the CORR values increased by 0.63%, 1.05%, and 0.37%, respectively. Thus, the prediction accuracy was effectively improved.

show abstract

“…Will the number of days in the sliding window affect the performance of deep learning? This research uses TPA-LSTM [4], Prophet, ARIMA [3] to conduct module testing. Furthermore, put forward the model test of CNN BiLSTM Attention, CNN BiGRU Attention, CNN BiGRU Dual Attention CNN BiLSTM Dual Attention.…”

Section: Introductionmentioning

confidence: 99%

Applying dual attention deep learning model to predict oil futures prices

Yen

Huang

2022

Preprint

View full text Add to dashboard Cite

Due to the non-linear and complex characteristics of oil futures prices, they are easily affected by external factors. Therefore, predicting the price of oil futures is a very challenging topic in deep learning time series, but the existing literature lacks research on introducing relevant features. Therefore, the purpose of this paper is to use a two-way attention mechanism to predict oil futures prices and import gold prices as features. The data source is taken from Yahoo Finance, providing 5146 items from 2000/02/28 to 2020/06/29. The paper uses CNNBiLSTM, CNNBiGRU models and Attention CNNBiLSTM, Attention CNNBiGRU, Dual Attention CNNBiGRU, Dual Attention CNNBiLSTM and TPA-LSTM, Prophet, ARIMA that increase the attention mechanism to conduct experiments to find the best model. Moreover, explore the sliding window review days to deal with two-dimensional time series data and the influence of the delay of the typical days to predict the experiment's impact. The experimental results show that increasing the number of review days will decrease the prediction accuracy. The new model is imported using the two-dimensional time series method, and the attention mechanism is integrated. Importing CNNBiGRU with a dual attention mechanism from 80% of the training items and 20% of the test items in the experimental data set can improve the overall prediction accuracy. For example, the model BiGRU has improved the RMSE from 2.79 to 1.85 by adding a dual attention mechanism. In addition, it is found that if the regression model has large fluctuations in the prediction data, it will increase the RMSE and MAE, especially in the experimental items of 96% training and 4% testing data sets.

show abstract

Discovery and Prediction of Stock Index Pattern via Three-Stage Architecture of TICC, TPA-LSTM and Multivariate LSTM-FCNs

Cited by 18 publications

References 40 publications

Stock index trend prediction based on TabNet feature selection and long short-term memory

Stock index trend prediction based on TabNet feature selection and long short-term memory

Swing Trend Prediction of Main Guide Bearing in Hydropower Units Based on MFS-DCGNN

Applying dual attention deep learning model to predict oil futures prices

Contact Info

Product

Resources

About