Stock prediction requires the joint modeling of temporal dynamics and cross-stock dependence. Existing graph-based and hypergraph-based forecasting methods often process spatial relation modeling and temporal evolution in separate stages, which may weaken the interaction between relational information and recurrent state updating. This study proposes a Recurrent Spatiotemporal Hypergraph Attention Gated Recurrent Unit model for stock forecasting, in which hypergraph-based higher order dependence and temporal dynamics are integrated within each recurrent update. The hypergraph is constructed offline from heterogeneous financial features through Tucker decomposition, similarity estimation, and Top-K sparsification, and is then used as a structured relational prior during forecasting. Experiments on CSI 300 constituent stocks from January 2014 to October 2024 show that RST-HGA-GRU achieves the best overall performance across multiple evaluation metrics and forecasting horizons from 1 to 6 days. Ablation, sensitivity, back testing, and multi-horizon Diebold–Mariano tests further support the effectiveness and robustness of the proposed framework. These results demonstrate that recurrent spatiotemporal fusion with hypergraph-based higher-order relation modeling is effective for stock price forecasting.

Time series forecasting is a critical task in various domains such as finance, energy, healthcare, and economics, playing a key role in supporting decision-making and strategic planning. However, traditional methods often struggle when dealing with nonlinear, noisy, and complex data structures. This paper proposes a novel forecasting model that integrates fuzzy clustering analysis with machine learning to leverage both hidden grouping patterns in the data and the flexible learning capability of algorithms. The input data were clustered, and fuzzy relationships were established based on fuzzy principles, thereby forming the fuzzy time series. Then, a K-Nearest Neighbors (KNN) model was employed to forecast the fuzzy-transformed time series, and performance metrics including MAE, MAPE, and MSE were computed on the predicted and tested sets to evaluate the model’s effectiveness. The proposed model is implemented in Python and validated using real-world datasets, particularly the TSC stock price series. Experimental results show that the proposed model significantly outperforms several conventional machine learning approaches in terms of forecasting accuracy, demonstrating its effectiveness and potential for practical applications in financial time series forecasting.

Stock price prediction is vital for quantitative investment but challenging due to multi-source data complexity, including endogenous, exogenous, and noise components. Standard deep learning models rely on end-to-end modeling of raw market data, failing to disentangle these distinct drivers and hindering prediction accuracy. To address this, we propose MMCAD-Net, a novel model based on time series decomposition. It first decomposes the original stock series into an exogenous cyclical component, endogenous temporal component and residual component, thereby disentangling the mixed temporal patterns. Subsequently, deep feature extraction and information refinement are applied to each component: multi-scale convolutions capture diverse patterns in the cyclical component; multi-level convolutional networks refine local and global features in the temporal component; and an attention mechanism sifts for potentially informative signals within the residuals. Finally, a multi-source feature aggregation mechanism fuses all enhanced information. Experiments on real-world stock market datasets demonstrate that MMCAD-Net surpasses mainstream models in both prediction accuracy and efficiency. Ablation studies further confirm the necessity and effectiveness of each core module.

Introduction: Accurate stock prediction faces significant hurdles due to market volatility and complex temporal dependencies. Existing deep learning models struggle to balance local feature extraction and global pattern capture. Methods: Introduce AttnFuse to model multi-scale financial time-series patterns while synergistically improving robustness. The architecture integrates two core components: an improved DA-RNN incorporating Multi-Head Attention and Residual Connections for refined feature selection, and a lightweight Transformer with time embeddings for efficient long-range dependency modeling. Predictions from both modules are dynamically weighted using an MLP with residual correction. Four stock datasets are used, with evaluation metrics including RMSE, MAE, MAPE, and R². The model is compared against benchmarks such as CNN, LSTM, DARNN, and CNN-LSTM. Results: The AttnFuse model achieves state-of-the-art performance across all four stock datasets, delivering the lowest error rates and highest predictive accuracy. It significantly outperforms benchmarks (CNN, LSTM, DA-RNN, CNN-LSTM) with optimal metrics including RMSE, MAE, MAPE, and R2 . Ablation studies confirm critical contributions of all components: removing Multi-Head Attention increases RMSE by 1.1–7.5%, disabling residual connections elevates MAE by 4.6–29.2%, replacing the lightweight Transformer degrades RMSE by 7.7–45.6%, and excluding dynamic fusion raises MAPE by 0.8–22.9%. Discussion: Experiments across four different stock indices show that AttnFuse consistently outperforms the benchmark model on RMSE, MAE, MAPE, and other metrics. The results confirm its effectiveness in capturing complex dependencies in highly volatile financial time series. Conclusion: AttnFuse establishes a new state-of-the-art for stock forecasting through synergistic attention-based fusion. Its dynamic weighting mechanism and computational efficiency provide a robust solution for volatile markets

Stock Price Forecasting of European Football Clubs Using an LSTM Model

Stock prices pose significant challenges in the field of financial forecasting, being influenced by market noise and nonlinear factors. A CNN-LSTM hybrid model, Transformer, multi-layer perceptron (MLP), and long short-term memory network (LSTM) are used in this study to forecast Apple Inc. (AAPL) stock prices over a five-year historical period with accuracy. The data is preprocessed using Min-Max normalization and a 60-day sliding window. The model is trained in the PyTorch framework and optimized using the mean squared error (MSE) loss function. The CNN-LSTM model outperforms the other models in terms of MAE, MSE, RMSE, and all four indices, according to the experimental data. Compared to the baseline model (MLP), the CNN-LSTM model reduces the MSE by approximately 46%. This study validates the CNN-LSTM hybrid model's superiority in time series analysis, providing an efficient tool for financial decision-making and laying a solid foundation for future research on integrating external factors (such as market sentiment and macroeconomic indicators).

Stock price prediction remains an intriguing task due to the high volatility and complex temporal dependencies present in financial time-series data. Accurate prediction of the highest stock price is particularly important for investors seeking to identify market peaks and optimize trading strategies. This study investigates the effectiveness of Long Short-Term Memory (LSTM) networks in forecasting DELL’s highest stock price by analyzing the impact of different activation functions. Historical stock price data from 2016 to 2024 were used, and several preprocessing techniques, including data normalization and chronological train-test splitting, were applied. The LSTM models were trained for 100 epochs and evaluated using Root Mean Square Error (RMSE), Mean Squared Error (MSE), and Mean Absolute Error (MAE). The main contribution of this research is a comparative analysis of the sensitivity of LSTM prediction performance to different activation functions, namely ReLU, ELU, Sigmoid, and Tanh, in the context of high-volatility financial time-series data. The experimental results show that the LSTM model using the ReLU activation function achieved the best performance, with an RMSE of 0.557942, MSE of 0.311300, and MAE of 0.338773, outperforming the other activation functions. These findings demonstrate that activation function selection significantly influences LSTM forecasting performance. The results provide practical insights for financial analysts and investors in selecting appropriate deep learning configurations for more reliable stock price prediction.

Stock price forecasting is a complex, stationary, and nonlinear time series estimation problem influenced by numerous factors. This complexity renders basic models inadequate for producing accurate forecasts of future stock prices.. Thus, precise price forecasting is crucial in this intricate and dynamic market where participants strive to make well-informed decisions to minimize losses and maximize profits. Motivated by this necessity, the present study aims to forecast future prices of selected indices constructed according to both Islamic and traditional criteria and to propose the most effective forecasting model for market participants. . Twelve indices- six traditional and six Islamic- were examined in this context using the ARIMA, XGBoost, LSTM, and Decision Tree methods. The investigation revealed that machine learning models outperformed the conventional approaches in terms of outcomes. The optimal parameters were then acquired and XGBoost-Decision Tree, LSTM-XGBoost, and LSTM-Decision Tree hybrid models were developed based on the results obtained. In this regard, 84 distinct models with seven algorithms- 1 conventional, 3 machine learning, and 3 hybrid models- with optimized hyperparameters were applied to the 12 indices. RMSE values were used to evaluate model performance. The LSTM-Decision Tree model was shown to be the greatest predictor for the BIST Participation 100 Index, while the LSTM-XGBoost model was the best predictor for all other indices.

Guessing stock rates for the future is one of the key components of financial deals market analysis, since they can provide information that acts like risk management and investing alternatives. Deep learning’s (DLs) strong processing capabilities and aptitude to understand the non-technical, non- linear interactions have been able to establish considerable progress in stock price forecast. The LSTM networks are good for the forecasting in the problems related to time series such as stock price forecast since they have the capability of taking into account lagged effects and avoiding prospects like gradients that vanish. After considering the above performance matrix, which are the following: accuracy, precision, recall, and the F1-score, it will be possible for the models to be evaluated completely. These criteria give a comprehensive evaluation of the predictive accuracy, the robustness, and the generalization ability of the algorithms. Mind the results to show you that traditional ML methods, which can provide prompt and simple interpretable results, the LSTM-based model makes a better job at modeling the complex temporal patterns found in stock price data. The relative research sets out how DL can raise stock price prediction accuracy thus it enriches the financial analysts and investors with new insights. Our suggested LSTM model delivers 15.78% more accurate results than Linear Regression and 6.91% more accurate results than Random Forest, thus, it convinces everyone about its highlevel performance in stock price prediction.

The capital market serves as a vital investment channel where stock prices exhibit dynamic fluctuations influenced by macroeconomic factors and market sentiments. This study estimates daily stock prices of PT Astra Agro Lestari Tbk (AALI), a leading palm oil company, using hybrid ARIMA-ARCH-GARCH models. Employing quantitative time series analysis, the population comprises all daily AALI stock prices from January 1, 2021, to June 30, 2025 (1,145 observations), sampled purposively via Investing.com data. Analysis techniques include ADF stationarity tests, ACF-PACF correlograms, AIC/SC/HQ model selection, ARCH-LM heteroskedasticity tests, and forecasting accuracy evaluation. Results identify ARIMA(1,1,1) as optimal for mean modeling and GARCH(2,1) for volatility, achieving 53% average forecasting accuracy for July 31-August 5, 2025. The hybrid model effectively captures price patterns despite external influences.

As stock market data becomes more transparent and accessible to the public, along with the application of advanced financial technologies, the demand for accurate financial market forecasts is rising. This growing need is essential for developing effective investment strategies. There are many econometric models applied to financial time series data forecasts, each based on its unique assumptions and subjected to different situations. This paper discusses the ARIMA model in econometrics for time series data forecasting, specifically its application in predicting Netflix's closing stock prices. It analyzes the methodology used in this case study, highlighting the limitations of the ARIMA model in dynamic forecasting and its challenges in modeling financial time series data. After replicating steps in the Netflix stock price forecast, it finds that although the ARIMA model performs well in the short term, or the one-step-ahead forecast, it lacks explanatory power in dynamic forecasts, especially in the case of dealing with financial time series data.

Abstract The rapid economic growth of recent years has led to a surge in stock market participation, necessitating the need for accurate stock price predictions to mitigate investment risks and maximize returns. However, the dynamic nature of stock prices and their intrinsic volatility pose significant challenges to traditional statistical and machine learning (ML) models, which often struggle with overfitting, poor robustness, and limited generalization. To address these challenges, this study introduces a novel framework: EvoBagNet, an evolutionary Bagging ensemble learning model specifically designed for robust and high-accuracy stock price prediction. EvoBagNet is a scalable and efficient ensemble framework combining an Extra tree-based model, categorical boosting (CatBoost), and Light Gradient Boosting Machine (LGBM) as part of a bagging ensemble technique to enhance predictive performance. The framework incorporates Complete Empirical Mode Decomposition (CEEMD) to decompose time series data into intrinsic mode functions (IMFs) across varying frequency spectra, allowing for a more granular analysis of temporal patterns. Hyper-parameter tuning is conducted using a fast, single-objective evolutionary algorithm designed to converge efficiently on optimal configurations for the ensemble model. The framework is evaluated on datasets from nine prominent IT sector companies, employing six rigorous evaluation metrics to comprehensively assess performance. Experimental results highlight EvoBagNet’s superior accuracy, robustness, and scalability, outperforming state-of-the-art models across diverse scenarios and datasets. EvoBagNet demonstrated exceptional prediction accuracy across all datasets, achieving performance scores of 97.0% ± 0.7, 98.3% ± 0.5, 97.3% ± 0.8, 97.4% ± 0.6, 97.0% ± 1.0, 98.6% ± 0.4, 98.8% ± 0.4, 91.7% ± 1.2, and 98.4% ± 0.3 for Tech Mahindra, Mindtree, Infosys, Wipro, TCS, Mphasis, L&T Tech, HCL, and Coforge, respectively. These results highlight EvoBagNet’s potential as a powerful tool for stock price forecasting, offering significant implications for informed investment strategies and financial decision-making.

Stock price prediction remains a complex challenge due to the dynamic and non-linear nature of financial markets, especially for banking stocks like PT Bank Negara Indonesia (Persero) Tbk (BBNI). This study aims to optimize BBNI stock price forecasting by integrating an automated Extract, Transform, Load (ETL) pipeline with the Long Short-Term Memory (LSTM) algorithm within a data engineering framework. Historical data from 2019 to 2025 were processed through a structured ETL sequence—including data cleaning, feature engineering, and MinMaxScaler normalization—to ensure high data quality. The dataset was partitioned into 80% for model training and 20% for testing to ensure rigorous evaluation. The results demonstrate that the systematic ETL approach significantly enhances model stability and predictive accuracy compared to conventional methods. The LSTM model effectively captured long-term temporal dependencies, providing reliable trend forecasts with an impressive test accuracy, achieving a Root Mean Squared Error (RMSE) of 0.0354. This research underscores that integrating robust data engineering practices with deep learning is essential for building resilient financial decision-support systems.

Quantum Machine Learning (QML) prediction of stock values is on the verge of changing the financial market and particularly enhancing HFT techniques. This is an attempt at quantum computing-machine learning hybrid aimed at enhancing the accuracy and efficiency of trading decisions. The historical machine learning-based stock price prediction models are incapable of processing large volumes of data, executing trades faster, and modeling complex market dynamics in real-time. The challenges result in ineffective trading decisions and trade lags at high frequencies. These problems are solved by HFT Optimization and Quantum Machine Learning. Because quantum computing capability is processing large amounts of data simultaneously, the proposed architecture enhances prediction accuracy and speed while reducing latency of decision-making. Quantum Neural Networks (QNNs) and quantum-optimized algorithms are useful to boost the modeling of market behavior. This technology will enhance stock price forecasting and optimization of trading strategies, which increases profit and minimize risk. Initial evidence indicates that quantum-based HFT systems are faster in execution speed and market flexibility than conventional techniques, which is essential to the future of automated trading.

Aim: Stock price prediction remains a highly challenging task due to the complex and nonlinear nature of financial time series data. While deep learning (DL) has shown promise in capturing these nonlinear patterns, its effectiveness is often hindered by the low signal-to-noise ratio inherent in market data. This study aims to enhance the stock predictive performance and trading outcomes by integrating Singular Spectrum Analysis (SSA) with deep learning models for stock price forecasting and strategy development on the Australian Securities Exchange (ASX)50 index. Method: The proposed framework begins by applying SSA to decompose raw stock price time series into interpretable components, effectively isolating meaningful trends and eliminating noise. The denoised sequences are then used to train a suite of deep learning architectures, including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and hybrid CNN-LSTM models. These models are evaluated based on their forecasting accuracy and the profitability of the trading strategies derived from their predictions. Results: Experimental results demonstrated that the SSA-DL framework significantly improved the prediction accuracy and trading performance compared to baseline DL models trained on raw data. The best-performing model, SSA-CNN-LSTM, achieved a Sharpe Ratio of 1.88 and a return on investment (ROI) of 67%, indicating robust risk-adjusted returns and effective exploitation of the underlying market conditions. Conclusions: The integration of Singular Spectrum Analysis with deep learning offers a powerful approach to stock price prediction in noisy financial environments. By denoising input data prior to model training, the SSA-DL framework enhanced signal clarity, improved forecast reliability, and enabled the construction of profitable trading strategies. These findings suggested a strong potential for SSA-based preprocessing in financial time series modeling.

Purpose This study aims to explore the predictive role of artificial intelligence (AI)-driven sentiment analysis in financial markets by developing a hybrid long short-term memory–Random Forest framework. It investigates whether the integration of generative sentiment signals with historical market data can enhance the accuracy and robustness of stock price forecasting and financial predictions across various industry sectors. Design/methodology/approach This research uses a multisource data set from 2019 to 2024, including stock price data from Yahoo Finance, macroeconomic indicators from Federal Reserve Economic Data and textual sentiment from Reddit, Twitter, Bloomberg and Reuters. Transformer-based natural language processing models, such as FinBERT, are used to quantify sentiment, which is then used as a predictive feature in machine learning models. Granger causality analysis and accuracy metrics are applied to evaluate sectoral variations in sentiment impact. Findings Empirical analysis reveals that social media sentiment Granger causes short-term stock movements in technology and finance sectors, with the hybrid model achieving 68.5% directional accuracy and a 22% reduction in prediction error compared to ARIMA models benchmarks. In contrast, sectors like healthcare and energy show minimal sensitivity to sentiment, underscoring the need for domain-specific strategies. This study also identifies ethical concerns related to sentiment manipulation, transparency and AI governance in financial contexts. Originality/value This research introduces a reproducible, cross-sectoral forecasting framework that bridges AI, sentiment analysis and finance. The proposed architecture offers practical forecasting enhancements and contributes to ethical discourse on AI use in high-stakes financial environments, with implications for regulators, analysts and portfolio managers.

In the context of modern economic challenges, the effective functioning of the stock market becomes crucial for the sustainable development of the Russian economy. Attracting investments to the real sector requires creating effective tools for market information analysis, which involves processing large volumes of heterogeneous data under conditions of high volatility and geopolitical instability. The aim of the research is to develop an algorithm for building a sparse neural network. This model automatically eliminates insignificant connections between neurons for predicting stock market dynamics. The proposed approachis based on the method of solving a single-point inverse problem with a minimization of the sum of absolute parameter values, which allows reducing the model’s dimensionality. The scientific novelty of the research comprises two aspects. First, the work explores the possibility of using new factors generated by a large language model (an artificial intelligence system for text processing) for predicting stock market dynamics. Second, an original algorithm for constructing a sparse neural network has been developed. The research tested two main hypotheses. The first hypothesis aimed to verify the advantages of sparse neural networks over fully connected architectures in prediction accuracy. The second hypothesis investigated the effectiveness of using features extracted by large language models from unstructured text sources for financial forecasting.Experimental verification on three tasks of stock price and dividend forecasting confirmed both hypotheses. The sparse architecture demonstrated an advantage over fully connected models in prediction accuracy and computational efficiency. Automatic feature selection revealed the relevance of macroeconomic characteristics extracted by the large language model, confirming the promise of integrating modern natural language processing technologies into financial forecasting. The obtained results can be used to form effective strategies of stock market behavior and create intelligent decision support systems. In addition, sparse models can be used in solving other economic problems, including portfolio optimization andfinancial performance management.

Stock price prediction plays a significant role in supporting rational investment decision-making amidst the volatility of the Indonesian capital market. Accurately forecasting stock price movements, especially for leading stocks in the energy sector such as PT Aneka Tambang Tbk (ANTM), is crucial because these stocks play a key role in maintaining national economic stability. However, most previous research has been limited to linear models such as ARIMA, which are less able to capture non-linear and dynamic data patterns. This situation creates a research gap regarding the need for a more adaptive approach to the complexity of the stock market. To address this gap, this study offers a novel approach by applying an advanced machine learning approach based on Neural Networks (NN) to predict the stock price of PT Aneka Tambang Tbk (ANTM). The research data was obtained from the Investing.com website, covering the observation period from January 2020 to August 2025. The results showed that the Neural Network (NN) model was effective in predicting the weekly stock price of PT Aneka Tambang Tbk (ANTM), with the best performance achieved using the tanh activation function, an alpha value of 0.01, and a hidden layer architecture of 300;300. This model achieved high accuracy with an RMSE of 130.4853, an MAE of 91.5722, and a MAPE of 5.29%. These results indicate that the NN model successfully captures complex market patterns and provides accurate predictions, making it a valuable tool for investors and policymakers in making informed investment decisions.

ABSTRACT Forecasting high‐frequency stock prices is a significant challenge due to inherent noise, non‐stationarity, and complex market dynamics. Conventional models often struggle to effectively extract meaningful signals from raw price and volume data. To address this, we propose an innovative hybrid framework, EWOA‐VMD‐ATT‐BiGRU, which introduces several key novelties to enhance prediction accuracy. First, our model uniquely decomposes both price and volume sequences as an innovative feature engineering, which allows for the unveiling of multi‐scale market characteristics and effectively mitigates signal interference. Second, we employ an enhanced whale optimization algorithm (EWOA) to adaptively optimize the variational mode decomposition (VMD) parameters, ensuring a more precise and data‐driven signal separation. Finally, a Bidirectional GRU network integrated with an Attention mechanism (ATT‐BiGRU) is utilized to dynamically weigh the importance of the decomposed features for superior prediction. Empirical results on the high‐frequency SSE index demonstrate our model's superior performance. Compared to the second‐ranked model, it achieves reductions in MSE, RMSE, MAE, MSLE, MAPE, and SMAPE by 6.23%, 3.17%, 3.22%, 5.87%, 3.16%, and 3.15%, respectively. Notably, our strategy of decomposing both price and volume yields a substantial improvement, reducing key error metrics by over 40% compared to an equivalent non‐decomposed model. Our proposed decomposition‐based hybrid model offers a more precise and robust approach for forecasting of high‐frequency stock prices.

ABSTRACT Stock price prediction is challenging due to the high volatility and complex nonlinear patterns in financial markets. Traditional time series forecasting methods often struggle to capture such intricate dynamics. To address this, we propose a novel adaptive model fusion framework that integrates Convolutional Neural Networks (CNN), Bidirectional Long Short‐Term Memory networks (BiLSTM), and Attention Mechanisms (ATTN) to improve stock price forecasting. Our approach employs multi‐scale feature extraction using multiple CNNs enhanced with attention to adaptively select relevant features across different time horizons. An adaptive fusion mechanism dynamically adjusts the contribution of each sub‐model according to input data, optimizing predictions under varying market conditions. We investigated the effects of both univariate and multivariate data on model performance, and analyze how data distribution characteristics influence forecasting accuracy. Experiments were conducted on stock data from the top 9 companies in the Nasdaq 100 index by market capitalization, validating the robustness and effectiveness of our method across different sectors. Results show that our model significantly outperformed traditional forecasting approaches, achieving higher accuracy and improved generalization in diverse market environments. This study offers a novel framework for stock price prediction and provides valuable insights into adaptive model integration for financial time series forecasting.