Search Results (14)

Floods are among the most devastating natural hazards, intensified by climate change and rapid urbanization. This study introduces a novel deep learning framework, Transfer Learning-Enhanced FEDformer (TLE-FEDformer), designed for accurate and temporally consistent flood inundation mapping. The framework integrates pre-trained Xception backbones for robust multi-sensor feature extraction from Sentinel-1 Synthetic Aperture Radar (SAR) and Sentinel-2 optical imagery, a cross-modal fusion module to align heterogeneous modalities, and the Frequency Enhanced Decomposed Transformer (FEDformer) for efficient frequency-domain temporal modeling. This architecture effectively captures long-range dependencies and flood dynamics including onset, peak, duration, and recession, while addressing challenges such as cloud contamination, speckle noise, and limited labeled data. Comprehensive experiments demonstrate superior performance, achieving an overall accuracy of 98.12%, an F1-score of 98.55%, and an Intersection over Union (IoU) of 97.38%, outperforming baselines including Convolutional Neural Networks, Capsule Networks, and transfer learning alone. Ablation studies validate the contributions of each component, while sensitivity analyses confirm robustness across hyperparameters. Uncertainty quantification via Monte Carlo dropout highlights high confidence in core flooded regions. Preliminary generalization tests on independent events yield IoU > 94%, indicating strong transferability. TLE-FEDformer advances operational flood monitoring by providing reliable, scalable, and temporally consistent mapping from multi-sensor remote sensing data. This approach offers significant potential for real-time disaster response, early warning systems, and damage assessment in flood-prone regions worldwide. Full article

The stock market serves as an important channel for investors to preserve and increase their assets and has attracted significant attention. However, stock price is affected by multiple factors and represents complex characteristics such as high volatility, nonlinearity, and non-stationarity, making accurate prediction highly challenging. To improve forecasting accuracy, this study proposes FT-iTransformer, a stock price prediction model based on time–frequency domain collaborative analysis. The model integrates a frequency domain feature extraction module and a multi-scale temporal convolution network module to comprehensively capture both time and frequency domain features, and then the extracted features are fused and input into iTransformer. It models the complex relationships among multiple variables through the self-attention mechanism, utilizes the feedforward network to capture temporal dependencies, and finally the prediction results are output through the projection layer. This study conducts both comparative and ablation experiments on six stock datasets to evaluate the proposed FT-iTransformer model. The results of comparative experiments show that, compared with seven mainstream baseline models, such as LSTM, Informer, and FEDformer, FT-iTransformer achieves superior performance on all evaluation metrics. Furthermore, the results of ablation experiments exhibit the contributions of each core module to the overall predictive performance, and confirming the validity of the model’s design. In summary, FT-iTransformer provides an effective framework for predicting stock price accurately. Full article

Accurate forecasting in heterogeneous spatiotemporal environments requires models that are both generalizable and interpretable, while also preserving cross-scale symmetry between temporal and spatial patterns. Existing deep learning approaches often struggle with limited adaptability to data-scarce regions and lack transparency in capturing cross-scale causal factors. To address these challenges, we propose a novel framework, Cross-Scale Symmetry-Aware Causal Spatiotemporal Modeling with Adaptive Fusion and Region-Knowledge Transfer, which integrates three key innovations. First, a Dynamic Spatio-Temporal Fusion Framework (DSTFF) leverages frequency-aware temporal transformations and adaptive graph attention to capture complex multi-scale dependencies, ensuring temporal–spatial symmetry in representation learning. Second, a Region-Knowledge Enhanced Transfer Learning (RKETL) mechanism distills knowledge across regions through teacher–student distillation, graph-based embeddings, and meta-learning initialization, thereby maintaining structural symmetry between data-rich and data-scarce regions. Third, a Multi-Granularity Causal Inference Prediction Module (MCIPM) uncovers cross-scale causal structures and supports counterfactual reasoning, providing causal symmetry across daily, weekly, and monthly horizons. Comprehensive experiments on multi-regional logistics datasets from China and the U.S. validate the effectiveness of our approach. Across six diverse Chinese regions, our method consistently outperforms state-of-the-art baselines (e.g., PatchTST, TimesNet, FEDformer), reducing MAE by 18.5% to 27.4%. On the U.S. Freight dataset, our model achieves significant performance gains with stable long-horizon accuracy, confirming its strong cross-domain generalization. Few-shot experiments further demonstrate that with only 5% of training data, our framework surpasses the best baseline trained with 20% data. Robustness analyses under input perturbations and uncertainty quantification show that the model maintains low error variance and produces well-calibrated prediction intervals. Furthermore, interpretability is concretely realized through MCIPM, which visualizes the learned causal graphs and quantifies each regional factor’s contribution to forecasting outcomes. This causal interpretability enables transparent understanding of how temporal spatial dynamics interact across scales, supporting actionable decision-making in logistics management and policy planning. Overall, this work contributes a unified spatiotemporal learning framework that leverages symmetry principles across scales and regions to enhance interpretability, transferability, and forecasting accuracy. Full article

Forecasting real-world time series in domains with strong event sensitivity and regional variability poses unique challenges, as predictive models must account for sudden disruptions, heterogeneous contextual factors, and structural differences across locations. In tackling these challenges, we draw on the concept of symmetry that refers to the balance and invariance patterns across temporal, multimodal, and structural dimensions, which help reveal consistent relationships and recurring patterns within complex systems. This study is based on two multimodal datasets covering 12 tourist regions and more than 3 years of records, ensuring robustness and practical relevance of the results. In many applications, such as monitoring economic indicators, assessing operational performance, or predicting demand patterns, short-term fluctuations are often triggered by discrete events, policy changes, or external incidents, which conventional statistical and deep learning approaches struggle to model effectively. To address these limitations, we propose an event-aware multimodal time-series forecasting framework with graph-based regional transfer built upon an enhanced PatchTST backbone. The framework unifies multimodal feature extraction, event-sensitive temporal reasoning, and graph-based structural adaptation. Unlike Informer, Autoformer, FEDformer, or PatchTST, our model explicitly addresses naive multimodal fusion, event-agnostic modeling, and weak cross-regional transfer by introducing an event-aware Multimodal Encoder, a Temporal Event Reasoner, and a Multiscale Graph Module. Experiments on diverse multi-region multimodal datasets demonstrate that our method achieves substantial improvements over eight state-of-the-art baselines in forecasting accuracy, event response modeling, and transfer efficiency. Specifically, our model achieves a 15.06% improvement in the event recovery index, a 15.1% reduction in MAE, and a 19.7% decrease in event response error compared to PatchTST, highlighting its empirical impact on tourism event economics forecasting. Full article

Accurate hour-ahead photovoltaic (PV) forecasts are essential for grid balancing, intraday trading, and renewable integration. While Transformer architectures have recently reshaped time series forecasting, their application to short-term PV prediction with calibrated uncertainty remains largely unexplored. This study provides a systematic benchmark of five Transformer variants (Autoformer, Informer, FEDformer, DLinear, and PatchTST) evaluated on a five-year, rooftop PV dataset (5 kW peak) against an unseen 12-month test set. All models are trained within a pipeline using a 48-h rolling input window with cyclical temporal encodings to ensure comparability. Beyond point forecasts, we introduce Adaptive Conformal Inference (ACI), a distribution-free and adaptive framework, to quantify uncertainty in real time. The results demonstrate that PatchTST, through its patch-based temporal tokenization, delivers superior accuracy (MAE = 0.194 kW, RMSE = 0.381 kW), outperforming both classical persistence and other Transformer baselines. When coupled with ACI, PatchTST achieves 86.2% empirical coverage with narrow intervals (0.62 kW mean width) and probabilistic scores (CRPS = 0.54; Winkler = 1.86) that strike a balance between sharpness and reliability. The findings establish that combining patch-based Transformers with adaptive conformal calibration provides a novel and viable route to risk-aware PV forecasting. Full article

In recent years, Transformer-based methods have shown full potential in power load forecasting problems. However, their computational cost is high, while it is difficult to capture the global characteristics of the time series. When the forecasting time length is long, the overall shift of the forecasting trend often occurs. Therefore, this paper proposes a gray relation analysis–frequency-enhanced decomposition transformer (GRA-FEDformer) method for forecasting power loads in power systems. Firstly, considering the impact of different weather factors on power loads, the correlation between various factors and power loads was analyzed using the GRA method to screen out the high-correlation factors as model inputs. Secondly, a frequency decomposition method for long short-time-scale components was utilized. Its combination with the transformer-based model can give the deep learning model an ability to simultaneously capture the fluctuating behavior of the short time scale and the overall trend of changes in the long time scale in power loads. The experimental results show that the proposed method had better forecasting performance than the other methods for a one-year dataset in a region of Morocco. In particular, the advantages of the proposed method were more obvious in the forecasting task with a longer forecasting length. Full article

In recent years, emerging digital microfluidic technology has shown great application potential in fields such as biology and medicine due to its simple structure, sample-saving properties, ease of integration, and wide range of manipulation. Currently, due to potential faults in chips during production and usage, as well as high safety requirements in their application domains, thorough testing of chips is essential. This study records data using a machine vision-based digital microfluidic driving control system. As chip usage frequency rises, device degradation introduces seasonal and trend patterns in droplet motion time data, complicating predictive modeling. This paper first employs the density-based spatial clustering of applications with noise (DBSCAN) clustering algorithm to analyze the droplet motion time data in digital microfluidic systems. Subsequently, principal component analysis (PCA) is applied for dimensionality reduction on the clustered data. Using the INFORMER model, we predict changes in droplet motion time and conduct correlation analysis, comparing results with traditional long short-term memory (LSTM), frequency-enhanced decomposed transformer (FEDformer), inverted transformer (iTransformer), INFORMER, and DBSCAN-INFORMER prediction models. Experimental results show that the DBSCAN-PCA-INFORMER model substantially outperforms LSTM and other benchmark models in prediction accuracy. It achieves an R² of 0.9864, an MSE of 3.1925, and an MAE of 1.3661, indicating an excellent fit between predicted and observed values.The results demonstrate that the DBSCAN-PCA-INFORMER model achieves higher prediction accuracy than traditional LSTM and other approaches, effectively identifying the health status of experimental devices and accurately predicting failure times, underscoring its efficacy and superiority. Full article

Accurate and reliable medium- and long-term load forecasting is crucial for the rational planning and operation of power systems. However, existing methods often struggle to accurately extract and capture long-term dependencies in load data, leading to poor predictive accuracy. Therefore, this paper proposes a medium- and long-term residential load forecasting method based on FEDformer, aiming to capture long-term temporal dependencies of load data in the frequency domain while considering factors such as electricity prices and temperature, ultimately improving the accuracy of medium- and long-term load forecasting. The proposed model employs Discrete Cosine Transform (DCT) for frequency domain transformation of time-series data to address the Gibbs phenomenon caused by the use of Discrete Fourier Transform (DFT) in FEDformer. Additionally, causal convolution and attention mechanisms are applied in the frequency domain to enhance the model’s capability to capture long-term dependencies. The model is evaluated using real-world load data from power systems, and experimental results demonstrate that the proposed model effectively learns the temporal and nonlinear characteristics of load data. Compared to other baseline models, DCTformer improves prediction accuracy by 37.5% in terms of MSE, 26.9% in terms of MAE, and 26.24% in terms of RMSE. Full article

Due to its effectiveness as a risk-hedging trading strategy in financial markets, futures arbitrage is highly sought after by investors in turbulent market conditions. The essence of futures arbitrage lies in formulating strategies based on predictions of future futures price differentials. However, contemporary research predominantly focuses on projections of single indicators for the subsequent temporal juncture, and devising efficacious arbitrage strategies often necessitates the examination of multiple indicators across timeframes. To tackle the aforementioned challenge, our methodology leverages a PSO Deep-ConvLSTM network, which, through particle swarm optimization (PSO), refines hyperparameters, including layer architectures and learning rates, culminating in superior predictive performance. By analyzing temporal-spatial data within financial markets through ConvLSTM, the model captures intricate market patterns, performing better in forecasting than traditional models. Multistep forward simulation experiments and extensive ablation studies using future data from the Shanghai Futures Exchange in China validate the effectiveness of the integrated model. Compared with the gate recurrent unit (GRU), long short-term memory (LSTM), Transformer, and FEDformer, this model exhibits an average reduction of 39.8% in root mean squared error (RMSE), 42.5% in mean absolute error (MAE), 45.6% in mean absolute percentage error (MAPE), and an average increase of 1.96% in coefficient of determination (R2) values. Full article

Renewable energy production is one of the most important strategies to reduce the emission of greenhouse gases. However, wind and solar energy especially depend on time-varying properties of the environment, such as weather. Hence, for the control and stabilization of electricity grids, the accurate forecasting of energy production from renewable energy sources is essential. This study provides an empirical comparison of the forecasting accuracy of electricity generation from renewable energy sources by different deep learning methods, including five different Transformer-based forecasting models based on weather data. The models are compared with the long short-term memory (LSTM) and Autoregressive Integrated Moving Average (ARIMA) models as a baseline. The accuracy of these models is evaluated across diverse forecast periods, and the impact of utilizing selected weather data versus all available data on predictive performance is investigated. Distinct performance patterns emerge among the Transformer-based models, with Autoformer and FEDformer exhibiting suboptimal results for this task, especially when utilizing a comprehensive set of weather parameters. In contrast, the Informer model demonstrates superior predictive capabilities for onshore wind power and photovoltaic (PV) power production. The Informer model consistently performs well in predicting both onshore wind and PV energy. Notably, the LSTM model outperforms all other models across various categories. This research emphasizes the significance of selectively using weather parameters for improved performance compared to employing all parameters and a time reference. We show that the suitability and performance of a prediction model can vary significantly, depending on the specific forecasting task and the data that are provided to the model. Full article

Seasonal–trend-decomposed transformer has empowered long-term time series forecasting via capturing global temporal dependencies (e.g., period-based dependencies) in disentangled temporal patterns. However, existing methods design various auto-correlation or attention mechanisms in the seasonal view while ignoring the fine-grained temporal patterns in the trend view in the series decomposition component, which causes an information utilization bottleneck. To this end, a Transformer-based seasonal–trend decomposition methodology with a multi-scale attention mechanism in the trend view and a multi-view attention mechanism in the seasonal view is proposed, called DESTformer. Specifically, rather than utilizing the moving average operation in obtaining trend data, a frequency domain transform is first applied to extract seasonal (high-frequency) and trend (low-frequency) components, explicitly capturing different temporal patterns in both seasonal and trend views. For the trend component, a multi-scale attention mechanism is designed to capture fine-grained sub-trends under different receptive fields. For the seasonal component, instead of the frequency-only attention mechanism, a multi-view frequency domain (i.e., frequency, amplitude, and phase) attention mechanism is designed to enhance the ability to capture the complex periodic changes. Extensive experiments are conducted on six benchmark datasets covering five practical applications: energy, transportation, economics, weather, and disease. Compared to the state-of-the-art FEDformer, our model shows reduced MSE and MAE by averages of 6.5% and 3.7%, respectively. Such experimental results verify the effectiveness of our method and point out a new way towards handling trends and seasonal patterns in long-term time series forecasting tasks. Full article

Long series time forecasting has become a popular research direction in recent years, due to the ability to predict weather changes, traffic conditions and so on. This paper provides a comprehensive discussion of long series time forecasting techniques and their applications, using the Informer algorithm model as a framework. Specifically, we examine sequential time prediction models published in the last two years, including the tightly coupled convolutional transformer (TCCT) algorithm, Autoformer algorithm, FEDformer algorithm, Pyraformer algorithm, and Triformer algorithm. Researchers have made significant improvements to the attention mechanism and Informer algorithm model architecture in these different neural network models, resulting in recent approaches such as wavelet enhancement structure, auto-correlation mechanism, and depth decomposition architecture. In addition to the above, attention algorithms and many models show potential and possibility in mechanical vibration prediction. In recent state-of-the-art studies, researchers have used the Informer algorithm model as an experimental control, and it can be seen that the algorithm model itself has research value. The informer algorithm model performs relatively well on various data sets and has become a more typical algorithm model for time series forecasting, and its model value is worthy of in-depth exploration and research. This paper discusses the structures and innovations of five representative models, including Informer, and reviews the performance of different neural network structures. The advantages and disadvantages of each model are discussed and compared, and finally, the future research direction of long series time forecasting is discussed. Full article

The storage environment can significantly impact paddy quality, which is vital to human health. Changes in storage can cause growth of fungi that affects grain quality. This study analyzed grain storage monitoring data from over 20 regions and found that five factors are essential in predicting quality changes during storage. The study combined these factors with the FEDformer (Frequency Enhanced Decomposed Transformer for Long-term Series Forecasting) model and k-medoids algorithm to construct a paddy quality change prediction model and a grading evaluation model, which showed the highest accuracy and lowest error in predicting quality changes during paddy storage. The results emphasize the need for monitoring and controlling the storage environment to preserve grain quality and ensure food safety. Full article

Changes in storage environments have a significant impact on grain quality. Accurate prediction of any quality changes during grain storage in different environments is very important for human health. In this paper, we selected wheat and corn, which are among the three major staple grains, as the target grains whose storage monitoring data cover more than 20 regions, and constructed a grain storage process quality change prediction model, which includes a FEDformer-based grain storage process quality change prediction model and a K-means++-based grain storage process quality change grading evaluation model. We select six factors affecting grain quality as input to achieve effective prediction of grain quality. Then, evaluation indexes were defined in this study, and a grading evaluation model of grain storage process quality was constructed using clustering model with the index prediction results and current values. The experimental results showed that the grain storage process quality change prediction model had the highest prediction accuracy and the lowest prediction error compared with other models. Full article