Search

Using machine learning coupled with stock price data to predict stock price trends has attracted increasing attention from data mining and machine learning communities. An accurate prediction results can help investors reduce investment risks and improve investment returns. The research on correlation stocks is one of the most important directions among many studies. Due to the high volatility and randomness of stock data, the correlation between stocks changes over time, which makes the stock correlation in static correlation stock sets often inconsistent with reality. Furthermore, various raw data related to stocks contain sufficient stock history information to analyze the future trend of stocks, but traditional prediction models cannot make good use of this information, which restricts the learning ability of the model and reduces the prediction accuracy. In this paper, we propose a stock prediction model combining multi-view stock data features with dynamic market correlation information (MDF-DMC). The model extracts stock trend features by combining multi-view raw data of a single stock with a Multi-layer Perceptron Mixer (MLP-Mixer); The improved Transformer encoder learns the correlation between the stock to be predicted and all the selected stocks in the stock market dynamically and extracts the features of the market correlation. We have conducted a large number of experiments on a total of 578 stocks in the stock markets of China and the United States, and the results show that our model has achieved excellent accuracy and returns across all data sets. • Stock movements are influenced by both itself and the market. • Stock multidimensional data can be used to improve the forecasting performance. • The attention mechanism can effectively extract the correlation of market dynamics. [ABSTRACT FROM AUTHOR]

Accurate forecasting of the agricultural power load is important for rural electricity networks' safe and stable operation. But it is more uncertain and difficult to forecast than industrial, commercial, and residential loads. Multivariate time series prediction methods have problems such as poor refinement of weather services and difficulty in quantifying the influencing factors. Univariate time series prediction methods should be emphasized, and the characteristic information in the historical load data should be fully exploited. Therefore, this paper systematically investigates the decomposition strategy and attention-based recurrent neural network for univariate time-series multi-step prediction. The decomposition strategy is employed to generate derived variable sequences from the target sequence as inputs for the model, while the dual-stage attention mechanism is used for dynamically selecting features. A one-year real-world dataset of agricultural power loads in southern China is used for validation. The effectiveness of two different decomposition algorithms on dual-stage attention-based recurrent neural network (DARNN) is discussed, then the proposed ensemble empirical mode decomposition and dual-stage attention-based recurrent neural network (EEMD-DARNN) is compared with four univariate time series prediction models and three ablation models. The proposed model is optimal in four evaluation indicators, with R 2 , R M S E , M A P E , and M A E of 0.970, 12.298, 0.040, and 7.335 respectively in the shortest forecast horizon (15 min), and 0.941, 17.305, 0.065 and 11.347 respectively in the longest forecast horizon (1 h). It is a significant improvement over the existing models. In conclusion, EEMD-DARNN can achieve more accurate ultra-short-term agricultural power load forecasting. [ABSTRACT FROM AUTHOR]

The air pollutant diffusion trend prediction plays an important role in the environment protection. The approaches in the existing studies rarely consider spatial and temporal information sequentially or cannot use spatiotemporal structure information effectively. In this paper, we propose a new deep learning-based air pollutant diffusion trend prediction model, called GATBL-Learning. It is an end-to-end neural network model consisting of Graph Attention Network (GAT) and Encoder-Decoder model. Since each city in the region will set as a node, GAT can extract spatial information from them by aggregation operation. Then Encoder-Decoder model, consisting of Bi-LSTM as Encoder and LSTM as Decoder, mines time series features and outputs the data containing spatiotemporal features. In addition, we also observe that the changes of air pollutants show seasonality discipline and training the seasonal model can improve the model adaptability. The results demonstrate the proposed prediction model is effective and superior. [ABSTRACT FROM AUTHOR]

Stock price prediction and modeling demonstrate high economic value in the financial market. Due to the non-linearity and volatility of stock prices and the unique nature of financial transactions, it is essential for the prediction method to ensure high prediction performance and interpretability. However, existing methods fail to achieve both the two goals simultaneously. To fill this gap, this paper presents an interpretable intuitionistic fuzzy inference model, dubbed as IIFI. While retaining the prediction accuracy, the interpretable module in IIFI can automatically calculate the feature contribution based on the intuitionistic fuzzy set, which provides high interpretability of the model. Also, most of the existing training algorithms, such as LightGBM, XGBoost, DNN, Stacking, etc, can be embedded in the inference module of our proposed model and achieve better prediction results. The back-test experiment on China's A-share market shows that IIFI achieves superior performance — the stock profitability can be increased by more than 20% over the baseline methods. Meanwhile, interpretable results show that IIFI can effectively distinguish between important and redundant features via rating corresponding scores to each feature. As a byproduct of our interpretable methods, the scores over features can be used to further optimize the investment strategy. • An interpretable fuzzy inference model for stock prediction is proposed. • The proposed model is orthogonal to most of the existing training algorithm. • The contribution of features can be automatically calculated by the model. • The model can render interpretable information for stock prediction result. • The model achieves both high prediction performance and interpretability. [ABSTRACT FROM AUTHOR]

Natural gas, an efficient, eco-friendly and clean green energy, has become one of the important energy structures of various countries in the world, accurately predicting the production of natural gas can help the national energy agency solve "gas shortage" problem. To accurately predict natural gas production in China, this paper establishes an optimized grey system model with weighted fractional accumulation generation operation (abbreviated as WFNGBM(1,1,N)). The proposed model has all the advantages of the GMP(1,1,N) model, NGBM(1,1) model and weighted fractional accumulation generation operation, which makes it have excellent prediction performance. Moreover, five outstanding intelligent optimization algorithms (whale optimization algorithm, marine predators algorithm, grasshopper optimization algorithm, equilibrium optimization algorithm and arithmetic optimization algorithm) are used to solve the hyperparameters of the WFNGBM(1,1,N) model. It is found that the WFNGBM(1,1,N) model has the characteristics of convertibility and small sample modeling, which indicates that it is a small sample prediction model with strong compatibility. After confirming the feasibility of the proposed model compared with its competing models by using natural gas production in Germany, Italy and Canada as examples, the proposed model is used to study China's natural gas production. The results show that this model is very suitable for predicting and analyzing natural gas production in China. Based on this, the WFNGBM(1,1,N) model is used to estimate China's natural gas production in the next three years, and some reasonable suggestions are given according to the development trend of natural gas production. • The concept of weighted fractional accumulation generation operation is given. • An optimized nonlinear grey Bernoulli prediction model is established. • Five algorithms are used to solve the parameters of the proposed model. • The proposed model is used to study natural gas production in China. [ABSTRACT FROM AUTHOR]

• Novel grey forecasting model called the neural ordinary differential grey model (NODGM) is proposed. • NODGM can be learned through a training procedure. • NODGM has a wider application range and can learn the characteristics of different data samples. • The predicting sequence is obtained by using the Runge–Kutta method to solve the model. Due to the efficiency of grey models in predicting the time series of small samples, grey system theory has been well studied since it was first proposed and has now become an important method for small sample prediction. Inspired by the neural ordinary differential equations (NODE), this paper proposes a novel grey forecasting model called the neural ordinary differential grey model (NODGM). The NODGM model includes a novel whitening equation that allows the prediction model to be learned through a training procedure. Therefore, compared with other models whose structures and terms need to be artificially predefined based on the laws of real samples, the NODGM model has a wider application range and can learn the characteristics of different data samples. Then, to obtain a model with better prediction performance, we apply NODE to train the model. Finally, the predicting sequence is obtained by using the Runge–Kutta method to solve the model. In the experiments, we apply NODGM model to two energy samples and then contrast the experimental results with the results of some classical grey models to validate the effectiveness of NODGM model. The comparison results from predicting China's annual crude oil consumption and forecasting oilfield production in northern China show that the prediction accuracies achieved by NODGM model are 28% and 8% higher, respectively, than those achieved by the state-of-the-art grey forecasting models. [ABSTRACT FROM AUTHOR]