LSTM–Transformer-Based Robust Hybrid Deep Learning Model for Financial Time Series Forecasting

Financial Time Series Forecasting (TSF) remains a critical challenge in Artificial Intelligence (AI) due to the inherent complexity of financial data, characterized by strong non-linearity, dynamic non-stationarity, and multi-factor coupling. To address the performance limitations of Spiking Neural Networks (SNNs) caused by hyperparameter sensitivity, this study proposes an SNN model optimized by an Improved Cuckoo Search (ICS) algorithm (termed ICS-SNN). The ICS algorithm enhances global search capability through piecewise-mapping-based population initialization and introduces a dynamic discovery probability mechanism that adaptively increases with iteration rounds, thereby balancing exploration and exploitation. Applied to futures market price difference prediction, experimental results demonstrate that ICS-SNN achieves reductions of 13.82% in MAE, 21.27% in MSE, and 15.21% in MAPE, while improving the coefficient of determination (R2) from 0.9790 to 0.9822, compared to the baseline SNN. Furthermore, ICS-SNN significantly outperforms mainstream models such as Long Short-Term Memory (LSTM) and Backpropagation (BP) networks, reducing prediction errors by 10.8% (MAE) and 34.9% (MSE), respectively, without compromising computational efficiency. This work highlights that ICS-SNN provides a biologically plausible and computationally efficient framework for complex financial TSF, bridging the gap between neuromorphic principles and real-world financial analytics. The proposed method not only reduces manual intervention in hyperparameter tuning but also offers a scalable solution for high-frequency trading and multi-modal data fusion in future research.

This literature review provides a comprehensive overview of key developments in quantitative trading theory, machine learning-based financial time series forecasting, deep learning-based financial time series forecasting, and modern quantitative investment strategies. It highlights seminal contributions from renowned scholars and researchers in the field. This review first explores the Efficient Market Hypothesis (EMH) proposed by Eugene Fama in 1970 and its empirical validation, which lays the foundation for understanding stock market dynamics. It then focuses on the application of machine learning to financial time series forecasting, including Hulls Delta strategy, Hujll Js multifactor model regression series, and Junhua Chens seminal work in 2009, which emphasizes the role of machine learning in solving the challenges of financial time series data. Finally, an overview of the development of deep learning in financial time series forecasting is presented, including the comprehensive model of Bowie et al., the success of Junhua Chens Deep Belief Network (DBN), and the sequence data processing capabilities of the Multi-stage Attention Network (MAN) model. In terms of modern quantitative investment strategies, it covers research areas such as EBIT/EV-based stock ranking studies, higher-order moment analysis, frequent trading strategies, and investor sentiment indicators. In summary, this literature review showcases the evolution of quantitative trading theory, the emergence of machine learning and deep learning in financial time series forecasting, and the development of modern quantitative investment strategies, offering valuable insights and tools for investors, researchers, and practitioners in financial markets.

In this paper, we develop a method based on a deep learning method in financial market prediction, which includes BRICS economies as the test cases. Financial markets are rife with volatility that is affected by a "bed of complexity," coddled by local and distal factors. To leverage these vast datasets both deep learning models such as Convolutional Neural Networks (CNNs), Long Short Term Memory (LSTM) networks as well as hybrid architectures are used in this study. The paper evaluates the predictive accuracy of the models, and by so doing, identifies their strengths in predicting temporal dependencies and intricate market patterns. In particular, deep learning techniques are applied to case studies of individual countries in the BRICS to highlight the application of deep learning to disparate country specific problems, such as liquidity crises and market shocks. These findings show that classical statistical methods are outperformed by deep learning systems in a precise and reliable financial forecasting. This research highlights the ability of AI driven systems to change financial decision making processes, improving investor confidence and improving economic stability in BRICS nations. This study also readers the value of deep learning in financial market analysis, especially in economies in the developing countries. Application of techniques and architectures e.g. Convolutional Neural Networks (CNNs) that excel at identifying spatial patterns, and Long Short-Term Memory (LSTM) networks renowned for their prowess on sequential and time series data, for real world market prediction are explained. In addition, the research discusses hybrid architectures which extend knowledge, fusing strengths of both architectures to improve prediction accuracy and how deep learning develops to solve particular financial challenges. Through reading these notes readers get exposed to data preprocessing techniques such as normalization and feature selection which are important for boosting deep learning performance. The paper also includes an introduction to the evaluation of models using MSE and R-squared values for validating them in terms of reliable outputs. This research combines deep learning theory and practical case study to offer a useful educational resource for students, researchers, and practitioners who want to apply AI in financial forecasting in complex and dynamic global markets.

This article proposes a lightweight deep learning model based on residual thinking, RAR-LSTM (Residual and Regime aware LSTM), to address the difficult issues of high noise, non stationarity, and state switching in economic and financial time series forecasting. This method uses interpretable naive models as baselines (such as seasonal naive method for quarterly forecasting and random walk method for financial daily frequency forecasting), and only learns the residual part through small-scale LSTM, thus balancing stability and reproducibility. At the same time, introducing minimalist state features (such as quarterly dummy variables, high and low rolling volatility states) helps the model perceive changes in the macro or market environment; In terms of loss function, Pinball Loss is supported for quantile prediction to meet the demand for asymmetric underestimation and overestimation costs in practical business. This article validates the generality of RAR-LSTM in two typical tasks: the macro quarterly task is to predict the US unemployment rate, and the financial daily frequency task is to predict the volatility of S&P 500 days. The experimental results show that RAR-LSTM significantly outperforms methods such as naive baseline, traditional LSTM, and SARIMA in terms of prediction accuracy (MAE, RMSE, etc.), especially in high volatility states where it performs more robustly. This model has the characteristics of simple structure, low computational cost, and easy reproducibility, and is suitable for economic and financial forecasting tasks in teaching, competition, and practical scenarios. Beyond naive baselines, we benchmark against stronger statistical models (ARIMA and ETS) and show that RAR-LSTM remains robust across baselines. We also report quantile forecasts (τ=0.1/0.5/0.9) using Pinball Loss, demonstrating calibrated coverage and practical benefits for risk-aware decisions.

Financial time series forecasting model based on CEEMDAN and LSTM

Abstract—Stock price prediction remains one of the most challenging tasks in financial forecasting due to the inherent volatility and nonlinearity of financial markets. This paper presents a comprehensive review of machine learning approaches for stock price prediction, comparing traditional statistical methods with modern deep learning architectures. We analyze various algorithms including Support Vector Machines (SVM), Random Forest, Artificial Neural Networks (ANN), and Long Short-Term Memory (LSTM) networks. Our comparative analysis reveals that LSTM-based deep learning models achieve superior performance with prediction accuracies ranging from 85% to 89%, significantly outperforming traditional regression methods. The study examines key technical indicators, feature engineering techniques, and model optimization strategies that enhance prediction accuracy. We also discuss the challenges of market volatility, overfitting, and computational complexity. The findings demonstrate that sophisticated neural network architectures, particularly LSTM networks, offer more reliable predictions for investors and financial analysts in volatile market conditions. This research contributes to the growing body of knowledge on intelligent financial forecasting systems and provides practical insights for implementing machine learning-based trading strategies. Index Terms—Stock price prediction, machine learning, LSTM, deep learning, financial forecasting, technical analysis, time series prediction

In today's digital world, app stores have become an essential part of software distribution, providing customers with a wide range of applications and opportunities for software developers to showcase their work. This study elaborates on the importance of end-user feedback for software evolution. However, in the literature, more emphasis has been given to high-rating & popular software apps while ignoring comparatively low-rating apps. Therefore, the proposed approach focuses on end-user reviews collected from 64 low-rated apps representing 14 categories in the Amazon App Store. We critically analyze feedback from low-rating apps and developed a grounded theory to identify various concepts important for software evolution and improving its quality including user interface (UI) and user experience (UX), functionality and features, compatibility and device-specific, performance and stability, customer support and responsiveness and security and privacy issues. Then, using a grounded theory and content analysis approach, a novel research dataset is curated to evaluate the performance of baseline machine learning (ML), and state-of-the-art deep learning (DL) algorithms in automatically classifying end-user feedback into frequently occurring issues. Various natural language processing and feature engineering techniques are utilized for improving and optimizing the performance of ML and DL classifiers. Also, an experimental study comparing various ML and DL algorithms, including multinomial naive Bayes (MNB), logistic regression (LR), random forest (RF), multi-layer perception (MLP), k-nearest neighbors (KNN), AdaBoost, Voting, convolutional neural network (CNN), long short-term memory (LSTM), bidirectional long short term memory (BiLSTM), gated recurrent unit (GRU), bidirectional gated recurrent unit (BiGRU), and recurrent neural network (RNN) classifiers, achieved satisfactory results in classifying end-user feedback to commonly occurring issues. Whereas, MLP, RF, BiGRU, GRU, CNN, LSTM, and Classifiers achieved average accuracies of 94%, 94%, 92%, 91%, 90%, 89%, and 89%, respectively. We employed the SHAP approach to identify the critical features associated with each issue type to enhance the explainability of the classifiers. This research sheds light on areas needing improvement in low-rated apps and opens up new avenues for developers to improve software quality based on user feedback.

Accurate financial time series forecasting is essential for risk management, portfolio optimization, and trading strategies. Traditional statistical models and classical machine learning approaches often struggle to capture the complex dependencies and volatility of financial markets. Recent advancements in deep learning, particularly transformer-based architectures, have shown significant promise in modeling sequential financial data. However, integrating dynamic risk assessment into forecasting models remains an open challenge. This study proposes a transformer-based financial time series forecasting framework that incorporates adaptive risk metrics to improve predictive accuracy and risk-aware decision-making. The model leverages self-attention mechanisms to capture long-range dependencies in financial data while integrating dynamic volatility measures, value-at-risk (VaR), and conditional value-at-risk (CVaR) as additional input features. By incorporating these adaptive risk factors, the model enhances its ability to anticipate market fluctuations and adjust forecasts accordingly. Experiments on real-world financial datasets demonstrate that the proposed approach outperforms traditional autoregressive models, recurrent neural networks (RNNs), and baseline transformer architectures in terms of predictive accuracy and risk-adjusted performance. The results highlight the importance of integrating risk-sensitive metrics into deep learning-based financial forecasting models, offering a more comprehensive approach for market analysis and investment decision-making.

Monaural Speech Enhancement Using a Multi-Branch Temporal Convolutional Network

Echo state networks (ESNs), as efficient and powerful computational models for approximating nonlinear dynamical systems, have been successfully applied in financial time series forecasting. Reservoir constructions in standard ESNs rely on trials and errors in real applications due to a series of randomized model building stages. A novel form of ESN with deterministically constructed reservoir is competitive with standard ESN by minimal complexity and possibility of optimizations for ESN specifications. In this paper, forecasting performances of deterministic ESNs are investigated in stock price prediction applications. The experiment results on two benchmark datasets (Shanghai Composite Index and S&P500) demonstrate that deterministic ESNs outperform standard ESN in both accuracy and efficiency, which indicate the prospect of deterministic ESNs for financial prediction.

In the dynamic landscape of financial markets, accurate forecasting of stock indices remains a pivotal yet challenging task, essential for investors and policymakers alike. This study is motivated by the need to enhance the precision of predicting the Shanghai Composite Index's opening price spread, a critical measure reflecting market volatility and investor sentiment. Traditional time series models like ARIMA have shown limitations in capturing the complex, nonlinear patterns inherent in stock price movements, prompting the exploration of advanced methodologies. The aim of this research is to bridge the gap in forecasting accuracy by developing a hybrid model that integrates the strengths of ARIMA with deep learning techniques, specifically Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. This novel approach leverages the ARIMA model's proficiency in linear trend analysis and the deep learning models' capability in modeling nonlinear dependencies, aiming to provide a comprehensive tool for market prediction. Utilizing a comprehensive dataset covering the period from December 20, 1990, to June 2, 2023, the study develops and assesses the efficacy of ARIMA, LSTM, GRU, ARIMA-LSTM, and ARIMA-GRU models in forecasting the Shanghai Composite Index's opening price spread. The evaluation of these models is based on key statistical metrics, including Mean Squared Error (MSE) and Mean Absolute Error (MAE), to gauge their predictive accuracy. The findings indicate that the hybrid models, ARIMA-LSTM and ARIMA-GRU, perform better in forecasting the opening price spread of the Shanghai Composite Index than their standalone counterparts. This outcome suggests that combining traditional statistical methods with advanced deep learning algorithms can enhance stock market prediction. The research contributes to the field by providing evidence of the potential benefits of integrating different modeling approaches for financial forecasting, offering insights that could inform investment strategies and financial decision-making.

This study aimed to improve apple slices’ color and drying kinetics by optimizing the hot-air drying process, utilizing machine and deep learning models. Different steam blanching times (30, 60, 90, and 120 s), drying temperatures (50, 55, 60, 65, and 70 °C), and humidity control methods (full humidity removal or temperature–humidity control) were examined. These factors significantly affected the quality of apple slices. 60 s blanching, 60 °C temperature, and full dehumidification represented the optimal drying conditions for apple slices’ dehydration, achieving better drying kinetics and the best color quality. However, the fastest drying process (40 min) was obtained at a 60 °C drying temperature combined with complete dehumidification after 90 s blanching. Furthermore, machine and deep learning models, including backpropagation (BP), convolutional neural network–long short-term memory (CNN-LSTM), temporal convolutional network (TCN), and long short-term memory (LSTM) networks, effectively predicted the moisture content and color variation in apple slices. Among these, LSTM networks demonstrated exceptional predictive performance with an R2 value exceeding 0.98, indicating superior accuracy. This study provides a scientific foundation for optimizing the drying process of apple slices and illustrates the potential application of deep learning in the agricultural processing and engineering fields.

The long short-term memory (LSTM) model is widely used in multiple areas, mainly for speech recognition, natural language processing and activity recognition. In the last few years, we started to see many variants of LSTM for recurrent neural networks since its inception in 1997. However, there weren’t many studies that have addressed the LSTM’s gating mechanism. In this paper, we propose a novel LSTM framework where we modify the architecture of the LSTM unit by adding a new layer that we call the “outlier gate”. The latter controls the flow of information that goes into the LSTM cell. This added signal allows us to avoid both the carry-over effect that the outliers have on the forecasted point and a bias in the estimates of our LSTM model – caused by unusual or non-repetitive events. The proposed architecture led us to an end-to-end trainable model that we applied in this paper to a financial time-series forecasting problem. Our results demonstrate that the new proposed LSTM architecture achieves better performance than the state-of-the-art original LSTM model.

In this paper, a time series forecasting approach by integrating particle swarm optimization (PSO) and support vector regression (SVR) is proposed. SVR has been widely applied in time series predictions. However, no general guidelines are available to choose the free parameters of an SVR model. The proposed approach uses PSO to search the optimal parameters for model selections in the hope of improving the performance of SVR. In order to evaluate the performance of the proposed approach, the TAIEX (Taiwan Stock Exchange Capitalization Weighted Stock Index) closing cash index is used as the illustrative example. Experimental results show that the proposed model outperforms the traditional SVR model and provides an alternative in financial time series forecasting.

Hydrate plugging and pipeline leak can impair the normal operation of natural gas pipeline and may lead to serious accidents. Since natural gas pipeline safety monitoring based on active acoustic excitation can detect and locate not only the two abnormal events but also normal components such as valves and pipeline elbows, recognition and classification of these events are of great importance to provide maintenance guidance for the pipeline operators and avoid false alarm. In this article, long short-term memory (LSTM) network is introduced and applied to classify detection signals of hydrate plugging, pipeline leak, and elbow. Adaptive moment estimation (Adam) algorithm is introduced and utilized to accelerate the long short-term memory network convergence in training. Experimental results demonstrate that the network with three layers and 64 units per cell performs the best. The cross-entropy loss in training is 0.0005, and classification accuracies are all 100% in training, validation, and testing which verify the validity of the long short-term memory network. Therefore, the method based on the long short-term memory network and adaptive moment estimation algorithm can work efficiently on pipeline events classification and has great guiding significance for safety assurance of natural gas transmission.