Hybrid Architectures that Combine LLMs and Predictive Analytics for Next-Generation Financial Modeling

Shiyang Chen; Shaochen Ren; Qun Zhang

doi:10.54097/j1cjb453

Authors

Shiyang Chen
Shaochen Ren
Qun Zhang

DOI:

https://doi.org/10.54097/j1cjb453

Keywords:

Large Language Models, Predictive Analytics, Hybrid Architectures, Financial Forecasting, Transformer Models, Multimodal Fusion, Sentiment Analysis, Explainable AI, Deep Learning, Portfolio Optimization

Abstract

The convergence of large language models (LLMs) and predictive analytics represents a transformative paradigm shift in financial modeling, offering unprecedented capabilities for processing multimodal data and generating actionable insights. This review examines the evolution, architecture, and applications of hybrid systems that integrate LLMs with traditional predictive models to address complex challenges in financial forecasting, risk management, and portfolio optimization. Recent advances in natural language processing (NLP) have enabled LLMs to extract nuanced sentiment and contextual information from vast textual datasets, while deep learning (DL) architectures such as long short-term memory (LSTM), gated recurrent units (GRU), and transformer-based models have demonstrated superior performance in capturing temporal dependencies within financial time series. The integration of these technologies through early, intermediate, and late fusion strategies has yielded hybrid architectures that leverage the complementary strengths of linguistic understanding and numerical prediction. This paper synthesizes current research on financial LLMs including BloombergGPT and FinGPT, explores attention mechanisms and multimodal data fusion techniques, and evaluates the application of these hybrid systems across sentiment analysis, stock prediction, portfolio management, and fraud detection. Critical challenges including explainability, regulatory compliance, computational efficiency, and data quality are examined alongside emerging solutions. The review concludes that hybrid architectures combining LLMs and predictive analytics represent the future of financial modeling, offering enhanced accuracy, interpretability, and adaptability to dynamic market conditions, while emphasizing the need for continued research in model transparency, ethical AI deployment, and standardized evaluation frameworks.

References

[1]Zhang, L., & Hua, L. (2025). Major issues in high-frequency financial data analysis: A survey of solutions. Mathematics, 13(3), 347.

[2]Li Y, Wang S, Pan W, et al. A survey of large language models for financial applications: Progress, prospects and challenges. arXiv preprint arXiv:2406.11903. 2024.

[3]Vallarino, D. (2025). An AI-Enhanced Forecasting Framework: Integrating LSTM and Transformer-Based Sentiment for Stock Price Prediction. Journal of Economic Analysis, 4(3), 1-15.

[4]Wu S, Irsoy O, Lu S, et al. BloombergGPT: A large language model for finance. arXiv preprint arXiv:2303.17564. 2023.

[5]Yang H, Liu XY, Wang CD. FinGPT: Open-source financial large language models. arXiv preprint arXiv:2306.06031. 2023.

[6]Lazarev A, Khvatov A. Utilizing modern large language models for financial trend analysis and digest creation. arXiv preprint arXiv:2510.01225. 2025.

[7]Sukma, N., & Namahoot, C. S. (2024). An algorithmic trading approach merging machine learning with multi-indicator strategies for optimal performance. IEEE Access.

[8]Aponte, R., Rossi, R. A., Guo, S., Dernoncourt, F., Yu, T., Chen, X., ... & Lipka, N. (2024). A Framework for Fine-Tuning LLMs using Heterogeneous Feedback. arXiv preprint arXiv:2408.02861.

[9]Wang, J., Liu, J., Zheng, W., & Ge, Y. (2025). Temporal heterogeneous graph contrastive learning for fraud detection in credit card transactions. IEEE Access.

[10]Addy, W. A., Ajayi-Nifise, A. O., Bello, B. G., Tula, S. T., Odeyemi, O., & Falaiye, T. (2024). AI in credit scoring: A comprehensive review of models and predictive analytics. Global Journal of Engineering and Technology Advances, 18(2), 118-129.

[11]Cambria E, Zhang Y, Mao R. A comprehensive review on financial explainable AI. Artificial Intelligence Review. 2024;57(4):1-42.

[12]Debnath, B., Wilson, S. K., Basu, S., Kompella, S. Y., Singha, R., Sahoo, S. K., ... & Kundu, A. (2025). The Pharmaceutical Industry's Future: How Artificial Intelligence is Transforming Medicine. Advanced Pharmaceutical Bulletin.

[13]Crisanto, J. C., Leuterio, C. B., Prenio, J., & Yong, J. (2024). FSI insights.

[14]Behera S, Kumar P, Dash R. Prediction based mean-value-at-risk portfolio optimization using machine learning regression algorithms. Engineering Applications of Artificial Intelligence. 2023;118:105271.

[15]Deng S, Zhu Y, Yu Y, Huang X. An integrated approach of ensemble learning methods for stock index prediction. Expert Systems with Applications. 2024;238:121710.

[16]Cheng D, Huang S, Wei F. Adapting large language models via reading comprehension. In: The Twelfth International Conference on Learning Representations. 2024.

[17]Kim AG, Muhn M, Nikolaev V. Financial statement analysis with large language models. arXiv preprint arXiv:2407.17866. 2024.

[18]Zhang, M. (2023). Strategic Management and Sustainability Transitions. Routledge eBooks (1st Edition, p. 10). Informa.

[19]Huang SC, Chan J, Huang P. FinBERT: A pre-trained financial language representation model for financial text mining. In: Proceedings of the 29th International Joint Conference on Artificial Intelligence. 2023. p. 4513-4519.

[20]Sako, K., Mpinda, B. N., & Rodrigues, P. C. (2022). Neural networks for financial time series forecasting. Entropy, 24(5), 657.

[21]Jing N, Wu Z, Wang H. Forecasting stock market indices using recurrent neural network based hybrid models: CNN-LSTM, GRU-CNN, and ensemble models. Applied Sciences. 2023;13(7):4644.

[22]Zhang Y, Chen K, Liu X. Data-driven stock forecasting models based on neural networks: A review. Expert Systems with Applications. 2024;237:122847.

[23]Lim B, Arık SÖ, Loeff N, Pfister T. Temporal fusion transformers for interpretable multi-horizon time series forecasting. International Journal of Forecasting. 2021;37(4):1748-1764.

[24]Zhou H, Zhang S, Peng J, et al. Informer: Beyond efficient transformer for long sequence time-series forecasting. In: Proceedings of the AAAI Conference on Artificial Intelligence. 2021;35(12):11106-11115.

[25]Wang Y, Zhang J, Chen L. A financial time-series prediction model based on multiplex attention and linear transformer structure. Applied Sciences. 2023;13(8):5175.

[26]Chen W, Li M, Zhang H. A systematic review for transformer-based long-term series forecasting. Artificial Intelligence Review. 2025;58(1):1-35.

[27]Wu, J., Wang, C., Zhang, Z., He, L., Chen, Y., & Cheng, W. (2025). Patterns in time series to image phase and multi-step forecasting of financial indices. Journal of King Saud University Computer and Information Sciences, 37(4), 59.

[28]Nguyen T, Park S, Kim J. Modality-aware transformer for financial time series forecasting. arXiv preprint arXiv:2310.01232. 2024.

[29]Gao J, Li P, Chen Z, Zhang J. A survey on deep learning for multimodal data fusion. Neural Computation. 2020;32(5):829-864.

[30]Wang, S., Zhu, L., Shi, L., Mo, H., & Tan, S. (2023). A survey of full-cycle cross-modal retrieval: From a representation learning perspective. Applied Sciences, 13(7), 4571.

[31]Jiao, T., Guo, C., Feng, X., Chen, Y., & Song, J. (2024). A Comprehensive Survey on Deep Learning Multi-Modal Fusion: Methods, Technologies and Applications. Computers, Materials & Continua, 80(1).

[32]Wang Z, Chen Y, Liu H. Cross-modal temporal fusion for financial market forecasting. arXiv preprint arXiv:2504.13522. 2024.

[33]Gramegna A, Giudici P. SHAP and LIME: An evaluation of discriminative power in credit risk. Frontiers in Artificial Intelligence. 2021;4:752558.

[34]Giudici P, Raffinetti E. SAFE artificial intelligence in finance. Finance Research Letters. 2023;56:104088.

[35]Ma, K., Zhang, J., Huang, X., & Wang, M. (2025). Leveraging transformer models to predict cognitive impairment: accuracy, efficiency, and interpretability. BMC Public Health, 25(1), 504.

[36]Arsenault, P. D., Wang, S., & Patenaude, J. M. (2025). A survey of explainable artificial intelligence (XAI) in financial time series forecasting. ACM Computing Surveys, 57(10), 1-37.

[37]Wang Z, Huang B, Tu S, Zhang K, Xu L. DeepTrader: A deep reinforcement learning approach for risk-return balanced portfolio management. In: Proceedings of the AAAI Conference on Artificial Intelligence. 2021;35(1):643-650.

[38]Jiang, Y., Olmo, J., & Atwi, M. (2024). Deep reinforcement learning for portfolio selection. Global Finance Journal, 62, 101016.

[39]Almahdi S, Yang SY. An adaptive portfolio trading system: Risk-return portfolio optimization using recurrent reinforcement learning. Expert Systems with Applications. 2023;207:267-279.

[40]Gigli, P. (2020). Tactical Asset Allocation and Machine Learning: Empirical Findings on Weights Portfolio Optimization with Elastic Net Regularization (Master's thesis, Universidade NOVA de Lisboa (Portugal)).

[41]Masuda J. Portfolio optimization using a hybrid machine learning stock selection model. MIT Master's Thesis. 2024.

[42]Sutiene, K., Schwendner, P., Sipos, C., Lorenzo, L., Mirchev, M., Lameski, P., ... & Cerneviciene, J. (2024). Enhancing portfolio management using artificial intelligence: literature review. Frontiers in artificial intelligence, 7, 1371502.

[43]Agrawal D, Gupta S, Mehta R. AI-enhanced portfolio management: Leveraging machine learning for optimized investment strategies. International Journal of Finance. 2024;15(3):112-134.

[44]Nejad FS, Ebadzadeh MM. Stock market forecasting using DRAGAN and feature matching. Expert Systems with Applications. 2024;244:122952.

[45]Yu, W., Kim, I. Y., & Mechefske, C. (2021). Analysis of different RNN autoencoder variants for time series classification and machine prognostics. Mechanical Systems and Signal Processing, 149, 107322.

[46]Cai L, Wang J, Peng J, et al. Multi-modal fusion in healthcare applications. Journal of Medical Systems. 2019;43(9):295.

[47]Wang, Y. (2021). Survey on deep multi-modal data analytics: Collaboration, rivalry, and fusion. ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM), 17(1s), 1-25.

[48]Rajagopal, S., Popat, K., Meva, D., Bajeja, S., & Mudholkar, P. (Eds.). (2025). Artificial Intelligence Based Smart and Secured Applications: Third International Conference, ASCIS 2024, Rajkot, India, October 16–18, 2024, Revised Selected Papers, Part V (Vol. 2428). Springer Nature.

[49]Krishna S, Tian Y, Xiong C. Modality dropout for multimodal learning. arXiv preprint arXiv:2310.15261. 2023.

[50]Kollamkudiyil, A. R., Simon, G. E., & George, J. A. (2025). Federated Medical Chatbot: Fine-Tuned with Llama-3.

[51]Joshi, S. (2025). Comprehensive Review of Artificial General Intelligence AGI and Agentic GenAI: Applications in Business and Finance. Available at SSRN 5250611.

[52]Zhou T, Niu P, Sun L, Jin R. One fits all: Power general time series analysis by pretrained language models. Advances in Neural Information Processing Systems. 2023;36:43322-43355.

[53]Lu, M., & Shen, Z. J. M. (2021). A review of robust operations management under model uncertainty. Production and Operations Management, 30(6), 1927-1943.

[54]Smith, K. M., & Chapman, M. P. (2023). On exponential utility and conditional value-at-risk as risk-averse performance criteria. IEEE Transactions on Control Systems Technology, 31(6), 2555-2570.

[55]Wang W, Chen Y, Zhao Z. Attentive statement fraud detection: Distinguishing multimodal financial data. Expert Systems with Applications. 2023;215:119384.

[56]Vallarino M. Forecasting stock prices with hybrid deep learning: LSTM-Transformer integration. Journal of Economic Analysis. 2025;4(3):109-123.

[57]Singh R, Kumar A. Risk-adjusted deep reinforcement learning for portfolio optimization: A multi-reward approach. International Journal of Computational Intelligence Systems. 2025;18(1):45-62.

[58]Ludkovski, M. (2023). Statistical machine learning for quantitative finance. Annual Review of Statistics and Its Application, 10(1), 271-295.

[59]Guo W, Yang Z, Wu S, Wang X, Chen F. Explainable enterprise credit rating using deep feature crossing. Expert Systems with Applications. 2023;220:119704.

[60]Zhan Y, Liu M, Chen K. Enhancing stock price prediction using GANs and transformer-based attention mechanisms. Empirical Economics. 2024;67(4):1523-1547.

[61]Jony, M. A. M., Arafat, M. S., Islam, R., Rafi, S. S., Jalil, M. S., & Hossen, F. (2024). AI-powered cybersecurity in financial institutions: Enhancing resilience against emerging digital threats. AIJMR-Advanced International Journal of Multidisciplinary Research, 2(6).

[62]Faccia, A., McDonald, J., & George, B. (2023). NLP sentiment analysis and accounting transparency: A new era of financial record keeping. Computers, 13(1), 5.

[63]Falade, P. V. (2023). Decoding the threat landscape: Chatgpt, fraudgpt, and wormgpt in social engineering attacks. arXiv preprint arXiv:2310.05595.

[64]Lopez-Lira A, Tang Y. Can ChatGPT forecast stock price movements? Return predictability and large language models. arXiv preprint arXiv:2304.07619. 2023.

[65]Patel R, Singh K, Kumar M. Churn prediction via multimodal fusion learning: Integrating customer financial literacy and behavioral data. arXiv preprint arXiv:2312.01301. 2023.

[66]Pippas, N., Ludvig, E. A., & Turkay, C. (2025). The Evolution of Reinforcement Learning in Quantitative Finance: A Survey. ACM Computing Surveys, 57(11), 1-51.

[67]Hassija V, Chamola V, Mahapatra A, et al. Interpreting black-box models: A review on explainable artificial intelligence. Cognitive Computing. 2024;16(1):45-74.

[68]Thompson, E. K., Buertey, S., & Kim, S. Y. (2025). How Important Is Corporate Social Responsibility for Corporate Financial Performance?: A Machine Learning Prediction and Model Interpretability Approach. Business Ethics, the Environment & Responsibility.

[69]Basu, S. (2025). Interpreting Deep Learning Models and Unlocking New Applications With It (Doctoral dissertation, University of Maryland, College Park).

[70]Olimid, A. P., Georgescu, C. M., & Olimid, D. A. (2024). Legal analysis of EU Artificial Intelligence Act (2024): Insights from personal data governance and health policy. Access to Just. E. Eur., 120.

[71]He, X. D., Kou, S., & Peng, X. (2022). Risk measures: robustness, elicitability, and backtesting. Annual Review of Statistics and Its Application, 9, 141-166.

[72]Buckmann, M., Joseph, A., & Robertson, H. (2023). An interpretable machine learning workflow with an application to economic forecasting. International Journal of Central Banking, 19(4), 449-552.

[73]Intelligence, H. C. A. (2024). Artificial Intelligence Index Report 2024: Public Data.

[74]Baron M, Brogaard J, Hagströmer B, Kirilenko A. Risk and return in high-frequency trading. Journal of Financial and Quantitative Analysis. 2019;54(3):993-1024.

[75]Hinton G, Vinyals O, Dean J. Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531. 2019.

[76]Catania L, Di Mari R, de Magistris PS. Dynamic discrete mixtures for high-frequency prices. Journal of Business and Economic Statistics. 2022;40(2):559-577.

[77]Sood S, Papasotiriou K, Vaiciulis M, Balch T. Deep reinforcement learning for optimal portfolio allocation: A comparative study. FinPlan Conference Proceedings. 2023;21:45-62.

[78]Gu J, Du W, Rahman AM, Wang G. Margin trader: A reinforcement learning framework for portfolio management. In: Proceedings of the Fourth ACM International Conference on AI in Finance. 2023. p. 610-618.

[79]Li Y, Du N, Song X, Yang X. Cardinality and bounding constrained portfolio optimization using safe reinforcement learning. In: 2024 International Joint Conference on Neural Networks. IEEE. 2024. p. 234-241.

[80]Vuković, D. B., Dekpo-Adza, S., & Matović, S. (2025). AI integration in financial services: a systematic review of trends and regulatory challenges. Humanities and Social Sciences Communications, 12(1), 1-29.

[81]Asta, L., Pisano, C., Sbrigata, A., Raffa, G. M., Scola, L., & Balistreri, C. R. (2025). Biomarkers in Heart Failure: A Review and a Wish. International Journal of Molecular Sciences, 26

[82]Wang J, Chen S, He Y, et al. FinBEN: A comprehensive benchmark for financial natural language processing. arXiv preprint arXiv:2402.12659. 2024.

[83]Zhou T, Ma Z, Wen Q, et al. FEDformer: Frequency enhanced decomposed transformer for long-term series forecasting. In: International Conference on Machine Learning. PMLR. 2022. p. 27268-27286.

[84]Abomakhelb, A., Jalil, K. A., Buja, A. G., Alhammadi, A., & Alenezi, A. M. (2025). A Comprehensive Review of Adversarial Attacks and Defense Strategies in Deep Neural Networks. Technologies, 13(5), 202.

[85]Chatzimparmpas, A., Martins, R. M., Jusufi, I., Kucher, K., Rossi, F., & Kerren, A. (2020, June). The state of the art in enhancing trust in machine learning models with the use of visualizations. In Computer Graphics Forum (Vol. 39, No. 3, pp. 713-756).

[86]Aggarwal, M., Khullar, V., & Goyal, N. (2024). A comprehensive review of federated learning: Methods, applications, and challenges in privacy-preserving collaborative model training. Applied Data Science and Smart Systems, 570-575.