indicator consensus modeling

Time-series forecasting suffers from poor generalization and low efficiency due to tight coupling among sequence representation, information extraction, and future projection. To address this, we propose a modular forecasting framework that decouples the pipeline into three independently optimizable stages—representation learning, information extraction, and target projection—enabling flexible, task-aware component configuration. Our approach innovatively integrates convolutional layers with a lightweight self-attention mechanism, achieving efficient local feature modeling while capturing long-range temporal dependencies. Evaluated on seven benchmark datasets, the method consistently outperforms existing state-of-the-art models in prediction accuracy, while requiring significantly fewer parameters and achieving faster training and inference speeds. This demonstrates substantial improvements in both statistical performance and computational efficiency.

This paper addresses the challenges of modeling complex temporal dependencies and poor market adaptability in stock price forecasting. Methodologically, it systematically compares LSTM, GRU, and attention-based models, and—novelty—integrates self-attention mechanisms into a multi-agent reinforcement learning (MARL) framework, coupled with real-time streaming data processing and dynamic policy optimization. Its key contributions are: (1) joint modeling of short- and long-term temporal dependencies via hybrid attention-MARL architecture; and (2) enhanced robustness to market regime shifts through multi-agent strategic interaction and adversarial coordination. Empirical evaluation demonstrates that the proposed model achieves a 12.6% improvement in prediction accuracy over single-model baselines; out-of-sample backtesting yields an annualized return of 23.4% and a Sharpe ratio of 2.1—substantially outperforming conventional time-series models.

This paper addresses the challenge of jointly modeling temporal technical indicators and static fundamental information—tasks poorly handled by single-model approaches. We propose a hybrid LSTM–Random Forest forecasting framework: an LSTM module captures deep sequential patterns from price time series, while a Random Forest integrates technical indicators (e.g., MACD, RSI) with macroeconomic and firm-level fundamentals; crucially, it incorporates a feature-importance-driven technical indicator selection mechanism. Evaluated on 10-day return prediction for international public companies, our method significantly outperforms baseline models—including standard LSTM, Random Forest, and XGBoost—in both predictive accuracy (p < 0.01) and out-of-sample Sharpe ratio. Results demonstrate that heterogeneous data fusion yields substantial, statistically robust gains in quantitative trading performance. The framework offers a novel, interpretable, and robust paradigm for intelligent trading powered by multi-source financial data.

To address challenges in long-term multivariate time-series forecasting—including difficulty in modeling uncertainty across both channel and temporal dimensions, inefficient information fusion, and the trade-off between accuracy and interpretability—this paper proposes a novel evidential multi-source information fusion paradigm grounded in Dempster–Shafer evidence theory. We introduce a pioneering dual-dimensional (channel- and time-aware) uncertainty-sensitive Basic Probability Assignment (BPA) module, integrating fuzzy-theory-driven BPA construction with a lightweight multi-source feature fusion network. Furthermore, we design a multi-source evidence fusion mechanism that ensures high interpretability, low computational complexity, and robustness to hyperparameter variations. Evaluated on multiple benchmark datasets, our method achieves state-of-the-art forecasting accuracy while reducing training time by 37% and model parameters by 52%. Under hyperparameter perturbations, the MAE variation remains below 1.2%.

This paper addresses poor model reproducibility and insufficient open-source implementations in time-series forecasting by proposing a lightweight, fully reproducible LSTM/GRU modeling paradigm. Methodologically, it constructs univariate sequence samples via sliding windows and evaluates performance using two metrics—RMSE and directional accuracy (DA)—on both synthetic activity data (Activities) and real-world financial data (BSE BANKEX). A key finding is that effective training requires only a single time series exhibiting repetitive patterns, without complex preprocessing or large-scale datasets. Experiments show that the proposed implementation significantly outperforms the “repeat last value” baseline for 1-step and 20-step predictions on Activities, while achieving comparable performance on BSE BANKEX. All code, datasets, and complete experimental configurations are publicly released to ensure full reproducibility and out-of-the-box usability.

This work addresses the limitations of existing foundation time series models, which suffer from high computational overhead, poor adaptability to dynamic data streams, and an inability to learn continuously—hindering their deployment in resource-constrained environments. To overcome these challenges, we propose TimeBlocks, a novel foundation modeling paradigm that uniquely integrates multi-task generalization, lightweight architecture, and continual calibration capabilities. TimeBlocks dynamically assembles compact models at inference time through a modular pool of model blocks and a routing strategy tailored to incoming data streams. Furthermore, it incorporates StreamCore, a streaming summarization algorithm that enables efficient online calibration. Extensive experiments demonstrate that TimeBlocks achieves significantly higher prediction accuracy than current methods across multiple datasets while maintaining low computational costs, enabling effective real-time forecasting under stringent resource constraints.

Pretrained time series foundation models often underperform on downstream tasks due to domain shift, task heterogeneity, scarce labeled data, and computational constraints. This work proposes the first systematic post-training framework, categorizing existing approaches along five dimensions based on their intervention points within the forecasting pipeline: parameter adaptation, context augmentation, model composition, output and uncertainty calibration, and compression with specialization. By delineating the design space and inherent limitations of each category, the framework offers a structured pathway to bridge the gap between pretraining and reliable deployment, thereby advancing the standardization and systematic development of time series post-training methodologies.

This work addresses the limitations of existing time series forecasting benchmarks, which often exhibit simplistic dynamics and suffer from data-dependent evaluation biases that can misrepresent model performance. To overcome these issues, we propose TimeSynth, the first structured synthetic framework that generates signals incorporating real-world characteristics such as non-stationarity, trends, periodicity, and phase modulation. By calibrating distributional shifts and noise levels based on parameters derived from real data, TimeSynth enables a systematic evaluation of diverse models—including linear methods, MLPs, CNNs, and Transformers—under complex temporal dynamics. Our experiments reveal that linear models degrade to simple oscillatory behavior under such complexity, whereas Transformers and CNNs demonstrate significantly superior adaptability and robustness.

This work addresses the lack of systematic understanding regarding the success and failure mechanisms of generative models on real-world sensor time-series data. The authors propose SensorGen, the first unified framework for generating and evaluating multi-domain, multimodal sensor signals. They conduct a comprehensive benchmark of five prominent generative model families—including flow matching, diffusion, and autoregressive models—across four domains, seven datasets, and twelve signal modalities, introducing novel techniques for time–frequency modeling and covariate integration. Their findings reveal that flow matching models consistently achieve the best overall performance, that signal characteristics substantially influence generation quality, and that high-fidelity synthetic data can significantly enhance downstream task performance, thereby demonstrating its practical utility.