Complex Time Series Data Research Articles

CEEMDAN-Based Hybrid Machine Learning Models for Time Series Forecasting Using MARS Algorithm and PSO-Optimization

Accurate prediction of time series data is crucial for informed decision-making and economic development. However, predicting noisy time series data is a challenging task due to their irregularity and complex trends. In the past, several attempts have been made to model complex time series data using both stochastic and machine learning techniques. This study proposed a CEEMDAN-based hybrid machine learning algorithm combined with stochastic models to capture the volatility of weekly potato price in major markets of India. The smooth decomposed component is predicted using stochastic models, while the coarser components, selected using MARS, are fitted into two different machine learning algorithms. The final predictions for the original series are obtained using optimization techniques such as PSO. The performance of the proposed algorithm is measured using various metrics, and it is found that the optimization-based combination of models outperforms the individual counterparts. Overall, this study presents a promising approach to predict price series using a hybrid model combining stochastic and machine learning techniques, with feature selection and optimization techniques for improved performance.

Application of the dynamic transformer model with well logging data for formation porosity prediction

Porosity, as a key parameter to describe the properties of rock reservoirs, is essential for evaluating the permeability and fluid migration performance of underground rocks. In order to overcome the limitations of traditional logging porosity interpretation methods in the face of geological complexity and nonlinear relationships, the Dynamic Transformer model in machine learning was introduced in this study, aiming to improve the accuracy and generalization ability of logging porosity prediction. Dynamic Transformer is a deep learning model based on the self-attention mechanism. Compared with traditional sequence models, Dynamic Transformer has a better ability to process time series data and is able to focus on different parts of the input sequence in different locations, so as to better capture global information and long-term dependencies. This is a significant advantage for logging tasks with complex geological structures and time series data. In addition, the model introduces Dynamic Convolution Kernels to increase the model coupling, so that the model can better understand the dependencies between different positions in the input sequence. The introduction of this module aims to enhance the model's ability to model long-distance dependence in sequences, thereby improving its performance. We trained the model on the well log dataset to ensure that it has good generalization ability. In addition, we comprehensively compare the performance of the Dynamic Transformer model with other traditional machine learning models to verify its superiority in logging porosity prediction. Through the analysis of experimental results, the Dynamic Transformer model shows good superiority in the task of logging porosity prediction. The introduction of this model will bring a new perspective to the development of logging technology and provide a more efficient and accurate tool for the field of geoscience.

A New and Effective Classification Method for Complex Time Series Based on Information Measure

The growing importance of time series information measure raises questions about how to effectively cluster a large number of nonlinear complex time series data and accurately extract more hidden information from them. In this paper, a clustering measurement and classification method for complex time series, the symmetrical exponential Tsallis relative information (SETRI) measure, is proposed, which aims to address these problems. The intrinsic characteristics of different types of time series information could be validly identified by this method. The modified multidimensional scaling (MDS) method, based on the SETRI measure, has the ability to display the data in the form of graphs for intuitive exhibition and accomplish the process of dimension reduction. The introduction of weighted permutation patterns allows a higher-accuracy classification not only for the time series dissimilarity quantification, but also for avoiding dispensable errors. Besides, the feasibility of the modified MDS classification method is visually and quantitatively verified by the simulated and real-world data. Compared with other MDS methods, the proposed method has better performance, which is reflected in the validity and rationality of the clustering results, thus further verifying the feasibility of the proposed method. Therefore, the new results will be helpful to develop complex data clustering and dimensionality reduction methods.

Hybrid convolutional long short‐term memory models for sales forecasting in retail

AbstractThis study proposes novel sales forecasting approaches that merge deep learning methods in a hybrid model. Long short‐term memory (LSTM) is adopted for modeling the temporal characteristics of the data, whereas the convolutional neural network (CNN) focuses on identifying and extracting relevant exogenous information. We propose stacked (S‐CNN‐LSTM) and parallel (P‐CNN‐LSTM) hybrid architectures to understand complex time series data with varying seasonal patterns and multiple products correlations. The performance drivers of both architectures were empirically tested with a real‐world multivariate retail dataset and outperformed when compared with simple neural network architectures and standard autoregressive methods for short and long‐term forecasting horizons. When compared with traditional predictive approaches, the proposed hybrid models reduce the computational complexity while providing flexibility and robustness.

DEEP RECURRENT NEURAL NETWORKS IN ENERGY DEMAND FORECASTING: A CASE STUDY OF KAZAKHSTAN'S ELECTRICAL CONSUMPTION

The critical transformation of the energy sector demands innovative approaches to ensure the reliability and efficiency of energy systems. In this pursuit, this study delved into the potential of Deep Recurrent Neural Networks (DRNNs) for forecasting energy demand, using a comprehensive dataset detailing Kazakhstan's electrical consumption over a span of two years. Traditional statistical models have historically played a role in energy demand prediction, but the growing intricacy of the energy landscape calls for more advanced solutions. The paper presented a comparison of the DRNN with other traditional and machine learning models and highlighted the superior performance of DRNNs, especially in capturing complex temporal relationships. The energy sector is confronting unprecedented challenges due to population growth and the integration of diverse energy sources, leading to increased demand and system strains. Accurate energy demand prediction is essential for system reliability. Traditional models, though widely used, often overlook intricate variables like weather patterns and temporal factors. Through rigorous methodology, encompassing exploratory data analysis, feature engineering, and hyperparameter optimization, an optimized DRNN model was developed. The results demonstrated the DRNN's exceptional capability in processing complex time-series data, as evidenced by its attainment of an R-squared value of 83.6%. Additionally, it achieved Mean Absolute Errors and Root Mean Squared Errors of less than 2%. However, there were noticeable deviations in some predictions, suggesting areas for refinement. This research underscores the significance of DRNNs in energy demand prediction, highlighting their advantages over traditional models while also noting the need for ongoing optimization. The findings underscore DRNN's promise as a robust forecasting tool, pivotal for the energy sector's future resilience and efficiency.

Development of technology predicting based on EEMD-GRU: An empirical study of aircraft assembly technology

Technology prediction has been the subject of many prior studies, which have the issues of the long-term effectiveness, the high uncertainty, and the low predictive accuracy. To address these problems, this study developed a model based on a mixed neural network that combines the Ensemble Empirical Mode Decomposition (EEMD) signal decomposition method with the Gated Recurrent Unit (GRU) deep learning model. In this study, the literature data is first preprocessed using Latent Dirichlet Allocation (LDA) topic modeling, and clusters of key technology topics are obtained accordingly. Secondly, within the identified technology topics, the EEMD signal processing method is employed to decompose complex time-series data into simpler subsequences, and GRU prediction models are established. Thirdly, the ultimate technological prediction results are obtained by integrating each subsequence's prediction results. In addition, Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) were used to evaluate the prediction results. Finally, the field of aircraft assembly technology is analyzed as a case study. The results show that the EEMD-GRU hybrid model excels in prediction accuracy, and brings a new perspective and method to the field of technological prediction.

Advanced Stock Market Forecasting: A Comparative Analysis of ARIMA-GARCH, LSTM, and Integrated Wavelet-LSTM Models

In the era of big data, accurate forecasting of corporate data is crucial for formulating effective strategies and decisions. This paper focuses on the prediction of key corporate indicators, taking TSLA, JD, MSFT, and TCEHY as case studies. It explores the application of three forecasting models: ARIMA-GARCH, LSTM, and Wavelet-LSTM. By comparing the predictive accuracy of these models, we find that each model has its strengths and weaknesses under different data characteristics. The study not only emphasizes the importance of accurate forecasting for corporate management and market prediction but also summarizes the adaptability and limitations of different models in dealing with complex time series data, providing valuable reference and insights for similar forecasting tasks.

Applications of Deep Learning Models for Forecasting and Modelling Rainwater in Moscow

To model and forecast complex time series data, machine learning has become a major field. This machine learning study examined Moscow rainfall data's future performance. The dataset is split into 65% training and 35% test sets to build and validate the model. We compared these deep learning models using the Root Mean Square Error (RMSE) statistic. The LSTM model outperforms the BILSTM and GRU models in this data series. These three models forecast similarly. This information could aid the creation of a complete Moscow weather forecast book. This material would benefit policymakers and scholars. We also believe this study can be used to apply machine learning to complex time series data, transcending statistical approaches.

Method of autoregression in application of singular-spectral analysis of time series for forecasting production of oil and gas industry products

More and more time series data are produced in various fields. It provides data for the research of time series analysis method, and promotes the development of time series research. Due to the generation of highly complex and large-scale time series data, the construction of forecasting models for time series data brings greater challenges. The theoretical aspects of using the model of singular-spectral analysis of time series with the use of autoregression are considered, and the justification of the expediency of using this model for forecasting the production of products for both the oil and gas industry and dual-use products is given. Both autoregressive model and decision tree model can be applied with the same degree of reliability for forecasting aggregate values of production.

Study on the Application of Improved Deep Convolutional Neural Network Algorithm in Broken Information Recovery

Abstract This study on the fusion of deep convolutional neural network (CNN) and extended short-term memory network (LSTM) aims to improve the efficiency and accuracy of broken information recovery. The challenges faced by traditional information recovery techniques are addressed through improved algorithms. The research methodology includes constructing CNN models to automatically extract features and combining LSTM networks to process complex time-series data. We conducted a detailed experimental evaluation of the CNN-LSTM fusion algorithm, including recovery of different types of corrupted data, and compared it with other algorithms. The results show that the CNN-LSTM fusion algorithm has the highest structural similarity (0.9545) and the most minor normalized mean square error (0.0016) for recovering broken video information, outperforming the methods using CNN or LSTM alone. The fusion algorithm dramatically reduces computation time and resource consumption for processing complex datasets. The combination of CNN and LSTM significantly improves the performance of broken information recovery, especially in processing video and audio data, and provides new ideas for future information processing techniques.

Predicting COVID-19 Trends with Comparative Analysis of ARIMA and ANN Models

This study uses Autoregressive Integrated Moving Average (ARIMA) and Artificial Neural Networks to predict COVID-19 in Pakistan. The pandemic epidemic that hit Wuhan, China, in December 2019 and affected millions of people worldwide. As of March 1, 2023, Pakistan had 296,149 confirmed cases, 6,298 deaths, and 280,970 recoveries. The predictive algorithms above will predict confirmed cases, deaths, and recoveries for 30 days. The approach collects time-series data on confirmed cases, deaths, and recoveries. The data is processed and analyzed using ARIMA and ANN models. These models were chosen because they can handle non-linear and complex time series data, making them excellent for pandemic prediction. The research hypothesis is that ARIMA and ANN models can accurately anticipate Pakistani COVID-19 case trends over the next 30 days. Correlation and MSE are used to compare models. Early results reveal that ARIMA and ANN models accurately estimate COVID-19 prevalence in Pakistan. An in-depth study of the methodology suggests adjustments to improve forecast accuracy. This study has significant ramifications. Accurate projections can help policymakers choose public health initiatives, saving lives and money. Successfully using these machine learning models could lead to their usage in epidemic prediction.

Optimizing coagulant dosage using deep learning models with large-scale data

Water treatment plants are facing challenges that necessitate transition to automated processes using advanced technologies. This study introduces a novel approach to optimize coagulant dosage in water treatment processes by employing a deep learning model. The study utilized minute-by-minute data monitored in real time over a span of five years, marking the first attempt in drinking water process modeling to leverage such a comprehensive dataset. The deep learning model integrates a one-dimensional convolutional neural network (Conv1D) and gated recurrent unit (GRU) to effectively extract features and model complex time-series data. Initially, the model predicted coagulant dosage and sedimentation basin turbidity, validated against a physicochemical model. Subsequently, the model optimized coagulant dosage in two ways: 1) maintaining sedimentation basin turbidity below the 1.0 NTU guideline, and 2) analyzing changes in sedimentation basin turbidity resulting from reduced coagulant dosage (5–20%). The findings of the study highlight the effectiveness of the deep learning model in optimizing coagulant dosage with substantial reductions in coagulant dosage (approximately 22% reduction and 21 million KRW/year). The results demonstrate the potential of deep learning models in enhancing the efficiency and cost-effectiveness of water treatment processes, ultimately facilitating process automation.

Multi-scale time series analysis using TT-ConvLSTM technique for bearing remaining useful life prediction

This study advocates the utilization of a parallel neural network (PNN) architecture for the estimation of remaining useful life (RUL) of rolling element bearings. Conventional machine learning methods have often fallen short in processing machinery sensor data efficiently and accurately, given its vast nature and the temporal dynamics of component health conditions. To address this limitation, a robust PNN architecture is proposed which incorporates multiple parallel processing pathways. These pathways are equipped with diverse input neurons, capturing data from mechanical component condition sensors. The output neuron then efficiently predicts the RUL, capturing the degradation stages throughout the component's lifespan. The PNN structure provides better accuracy and computation time by efficiently handling huge amounts of data simultaneously and integrating both spatial and temporal information present. Additionally, time-transformer and recurrent neural networks (RNNs) are used to handle complex time series data. Improvement methodologies, such as positional encoding combined with a multi-headed self-attention mechanism and the convolutional long short-term memory (ConvLSTM) neural network, are employed. These methodologies adeptly handle spatiotemporal dependencies inherent in multidimensional features extracted across the time domain, frequency domain, and time–frequency domain, thereby considerably boosting the model's efficacy. Two case studies are conducted on XJTU-SY rolling element bearing dataset and FEMTO bearing dataset to validate and display the generalization capabilities of the proposed methodology, where PNN performed exceptionally in terms of accuracy and efficiency. It is concluded that with a novel implementation of variational stride temporal window strategy and utilizing the full capabilities of a PNN, the RUL of bearings can be predicted very accurately. The algorithm strategy is very robust and can be applied to other machine components as well. Subsequently, the code for the algorithm will be open-sourced.

An Exponential Autoregressive Time Series Model for Complex Data

In this paper, an exponential autoregressive model for complex time series data is presented. As for estimating the parameters of this nonlinear model, a three-step procedure based on quantile methods is proposed. This quantile-based estimation technique has the benefit of being more robust compared to least/absolute squares. The performance of the introduced exponential autoregressive model is evaluated by means of four established goodness-of-fit criteria. The practical utility of the novel time series model is showcased through a comparative analysis involving simulation studies and real-world data illustrations.

Time-Series Neural Network: A High-Accuracy Time-Series Forecasting Method Based on Kernel Filter and Time Attention

This research introduces a novel high-accuracy time-series forecasting method, namely the Time Neural Network (TNN), which is based on a kernel filter and time attention mechanism. Taking into account the complex characteristics of time-series data, such as non-linearity, high dimensionality, and long-term dependence, the TNN model is designed and implemented. The key innovations of the TNN model lie in the incorporation of the time attention mechanism and kernel filter, allowing the model to allocate different weights to features at each time point, and extract high-level features from the time-series data, thereby improving the model’s predictive accuracy. Additionally, an adaptive weight generator is integrated into the model, enabling the model to automatically adjust weights based on input features. Mainstream time-series forecasting models such as Recurrent Neural Networks (RNNs) and Long Short-Term Memory Networks (LSTM) are employed as baseline models and comprehensive comparative experiments are conducted. The results indicate that the TNN model significantly outperforms the baseline models in both long-term and short-term prediction tasks. Specifically, the RMSE, MAE, and R2 reach 0.05, 0.23, and 0.95, respectively. Remarkably, even for complex time-series data that contain a large amount of noise, the TNN model still maintains a high prediction accuracy.

A novel hybrid approach to mooring tension prediction for semi-submersible offshore platforms

The accuracy of mooring tension prediction significantly affects the safety of semi-submersible offshore platform operations and the efficiency of production planning. Nevertheless, traditional single prediction models have difficulty accurately forecasting for complex measured mooring tension data. This work provides a hybrid method for predicting mooring tension on semi-submersible maritime platforms based on VMD, error correction, and the convolutional neural network–long short-term memory (CNN-LSTM) and convolutional neural network–bidirectional long short-term memory (CNN-BiLSTM) models. The proposed hybrid prediction model was trained, validated, and tested using monitored mooring tension data acquired from the “Deepsea One” semi-submersible ocean platform in the South China Sea under different orientations and sea conditions. The results showed that the proposed hybrid prediction model for mooring tension on semi-submersible offshore platforms had lower values of the root mean square error, mean absolute error, and mean absolute percentage error and higher values of the coefficient of determination (R2) than almost all of the other comparison models when predicting complex time series data of the actual monitored mooring tension. This indicated that the proposed hybrid prediction model had better precision and stability than other models. Furthermore, this method can be used to anticipate nonlinear, nonstationary time series data in different fields of maritime engineering.

Electrical Power Generation Forecasting from Renewable Energy Systems Using Artificial Intelligence Techniques

Renewable energy (RE) sources, such as wind, geothermal, bioenergy, and solar, have gained interest in developed regions. The rapid expansion of the economies in the Middle East requires massive increases in electricity production capacity, and currently fossil fuel reserves meet most of the power station demand. There is a considerable measure of unpredictability surrounding the locations of the concerned regions where RE can be used to generate electricity. This makes forecasting difficult for the investor to estimate future electricity production that could be generated in each area over the course of a specific period. Energy production forecasting with complex time-series data is a challenge. However, artificial neural networks (ANNs) are well suited for handling nonlinearity effectively. This research aims to investigate the various ANN models capable of providing reliable predictions for sustainable sources of power such as wind and solar. In addition to the ANN models, a state-of-the-art ensemble learning approach is used to improve the accuracy of predictions further. The proposed strategies can forecast RE generation accurately over short and long time frames, relying on historical data for precise predictions. This work proposes a new hybrid ensemble framework that strategically combines multiple complementary machine learning (ML) models to improve RE forecasting accuracy. The ensemble learning (EL) methodology outperforms long short-term memory (LSTM), light gradient boosting machine (LightGBM), and sequenced-GRU in predicting wind power (MAE: 0.782, MAPE: 0.702, RMSE: 0.833) and solar power (MAE: 1.082, MAPE: 0.921, RMSE: 1.055). It achieved an impressive R2 value of 0.9821, indicating its superior accuracy.

Novel complex-valued long short-term memory network integrating variational mode decomposition for soft sensor

Industrial process data is complex time series data including trend and periodicity. However, the existing soft sensor models only focus on the time series context. Therefore, a novel complex-valued long short-term memory (CVLSTM) integrating the variational mode decomposition (VMD) (CVLSTM-VMD) is proposed for the soft sensor. The VMD decomposes industrial process data into binary sequences containing trend and periodic signals, which are input into the bidirectional CVLSTM module to extract the trend and periodic features. Then, the global average pooling fuses trend and periodic features, and the complex-valued fully connected layer learns the dynamic relationship between the fused feature and the key indicator for the soft sensor. Finally, the CVLSTM-VMD is applied in the actual polypropylene process and the purified terephthalic acid (PTA) process. The experiment results show that the proposed CVLSTM-VMD can achieve 93% and 99% prediction accuracies in the polypropylene and the PTA datasets, respectively. Moreover, the CVLSTM-VMD can play a vital role in product quality control and new product development.

Enhanced Automated Deep Learning Application for Short-Term Load Forecasting

In recent times, the power sector has become a focal point of extensive scientific interest, driven by a convergence of factors, such as mounting global concerns surrounding climate change, the persistent increase in electricity prices within the wholesale energy market, and the surge in investments catalyzed by technological advancements across diverse sectors. These evolving challenges have necessitated the emergence of new imperatives aimed at effectively managing energy resources, ensuring grid stability, bolstering reliability, and making informed decisions. One area that has garnered particular attention is the accurate prediction of end-user electricity load, which has emerged as a critical facet in the pursuit of efficient energy management. To tackle this challenge, machine and deep learning models have emerged as popular and promising approaches, owing to their having remarkable effectiveness in handling complex time series data. In this paper, the development of an algorithmic model that leverages an automated process to provide highly accurate predictions of electricity load, specifically tailored for the island of Thira in Greece, is introduced. Through the implementation of an automated application, an array of deep learning forecasting models were meticulously crafted, encompassing the Multilayer Perceptron, Long Short-Term Memory (LSTM), One Dimensional Convolutional Neural Network (CNN-1D), hybrid CNN–LSTM, Temporal Convolutional Network (TCN), and an innovative hybrid model called the Convolutional LSTM Encoder–Decoder. Through evaluation of prediction accuracy, satisfactory performance across all the models considered was observed, with the proposed hybrid model showcasing the highest level of accuracy. These findings underscore the profound significance of employing deep learning techniques for precise forecasting of electricity demand, thereby offering valuable insights with which to tackle the multifaceted challenges encountered within the power sector. By adopting advanced forecasting methodologies, the electricity sector moves towards greater efficiency, resilience and sustainability.

Improved BIGRU Model and Its Application in Stock Price Forecasting

In order to obtain better prediction results, this paper combines improved complete ensemble EMD (ICEEMDAN) and the whale algorithm of multi-objective optimization (MOWOA) to improve the bidirectional gated recurrent unit (BIGRU), which makes full use of original complex stock price time series data and improves the hyperparameters of the BIGRU network. To address the problem that BIGRU cannot make full use of the stationary data, the original sequence data are processed using the ICEEMDAN decomposition algorithm to derive the non-stationary and stationary parts of the data and modeled with the BIGRU and the autoregressive integrated moving average model (ARIMA), respectively. The modeling process introduces a whale algorithm for multi-objective optimization, which improves the probability of finding the best combination of parameter vectors. The R2, MAPE, MSE, MAE, and RMSE values of the BIGRU algorithm, ICEEMDAN-BIGRU algorithm, MOWOA-BIGRU algorithm, and the improved algorithm were compared. An average improvement of 14.4% over the original algorithm’s goodness-of-fit value will greatly improve the accuracy of stock price predictions.