Accurate stock market prediction provides great opportunities for informed decision-making, yet existing methods struggle with financial data's non-linear, high-dimensional, and volatile characteristics. Advanced predictive models are needed to effectively address these complexities. This paper proposes a novel multi-layer hybrid multi-task learning (MTL) framework aimed at achieving more efficient stock market predictions. It involves a Transformer encoder to extract complex correspondences between various input features, a Bidirectional Gated Recurrent Unit (BiGRU) to capture long-term temporal relationships, and a Kolmogorov-Arnold Network (KAN) to enhance the learning process. Experimental evaluations indicate that the proposed learning structure achieves great performance, with an MAE as low as 1.078, a MAPE as low as 0.012, and an R2 as high as 0.98, when compared with other competitive networks.
KukretiVBhattC, et al.A stock market trends analysis of reliance using machine learning techniques. In: International conference on information systems and computer networks (ISCON), 2023, pp. 1–6. IEEE.
2.
ZhaoCWuMLiuJ, et al.Progress and prospects of data-driven stock price forecasting research. Int J Cognit Comput Eng2023; 4: 100–108.
3.
XuWFuZLiH, et al.A study of the impact of COVID-19 on the Chinese stock market based on a new textual multiple ARMA model. Stat Anal Data Min2023; 16: 5–15.
4.
AriyoAAAdewumiAO, et al. Stock price prediction using the ARIMA model. In: International conference on computer modelling and simulation, 2014, pp. 106–112. IEEE.
5.
HerwartzH. Stock return prediction under GARCH—an empirical assessment. Int J Forecasting2017; 33: 569–580.
6.
SusantiDLabittaKF, et al.Forecasting Indonesian stock Index using ARMA-GARCH model. Int J Quant Res Model2024; 5: 199–204.
7.
MaQ. Comparison of ARIMA, ANN and LSTM for stock price prediction. In: E3S web of conferences, 2020, Vol. 218, p. 01026. EDP Sciences.
8.
ChenYWuJWuZ. China's commercial bank stock price prediction using a novel K-means-LSTM hybrid approach. Expert Syst Appl2022; 202: 117370.
9.
KirisciMYolcuOC. A new CNN-based model for financial time series: TAIEX and FTSE stocks forecasting. Neural Process Lett2022; 54: 3357–3374.
10.
AgarwalSAlapattBP, et al.Stock price prediction using RNNs: a comparative analysis of RNN, LSTM, GRU, and BiRNN. In: International conference on sustainable computing and smart systems (ICSCSS), 2024, pp. 1074–1079. IEEE.
11.
WuJM-TLiZHerencsarN, et al.A graph-based CNN-LSTM stock price prediction algorithm with leading indicators. Multimedia Syst2023; 29: 1751–1770.
12.
YinTLiuCDingF, et al.Graph-based stock correlation and prediction for high-frequency trading systems. Pattern Recognit2022; 122: 108209.
13.
ShiYWangYQuY, et al.Integrated GCN-LSTM stock prices movement prediction based on knowledge-incorporated graphs construction. Int J Mach Learn Cybern2024; 15: 161–176.
14.
YañezCKristjanpollerW, et al.Stock market index prediction using transformer neural network models and frequency decomposition. Neural Comput Appl2024; 36: 15777–15797.
15.
BaoWCaoYYangY, et al.Data-driven stock forecasting models based on neural networks: a review. Inf Fusion2025; 113: 102616.
16.
YuanJJiangW, et al.Achieving more with less: a tensor-optimization-powered ensemble method. arXiv Preprint2024: arXiv:2408.02936.
17.
YuanJChenH, et al.A margin-maximizing fine-grained ensemble method. arXiv Preprint2024: arXiv:2409.12849.
18.
DongXYuZCaoW, et al.A survey on ensemble learning. Front Comput Sci2020; 14: 241–258.
19.
LiuFGuoS, et al. Application of an ANN and LSTM-based ensemble model for stock market prediction. In: 2024 IEEE 7th international conference on information systems and computer aided education (ICISCAE), 2024, pp. 390–395. IEEE.
20.
CorizzoRRosenJ. Stock market prediction with time series data and news headlines: a stacking ensemble approach. J Intell Inf Syst2024; 62: 27–56.
21.
DeviBUSundarDAlliP, et al.An effective time series analysis for stock trend prediction using ARIMA model for nifty midcap-50. Int J Data Min Knowledge Manage Process2013; 3: 65–78.
22.
KhanderwalSMohantyD. Stock price prediction using ARIMA model. Int J Mark Hum Resour Re- Search2021; 2: 98–107.
23.
MarisettyN. Prediction of popular global stock indexes volatility by using ARCH/GARCH models. GARCH Models (July 24, 2024), 2024.
24.
GuptaADhingraB. Stock market prediction using hidden markov models. In: 2012 Students conference on engineering and systems, 2012, pp. 1–4. IEEE.
25.
SuZYiB. Research on HMM-based efficient stock price prediction. Mobile Inf Syst2022; 2022: 8124149.
26.
MatteraROttoP. Network log-ARCH models for forecasting stock market volatility. Int J Forecasting2024; 40: 1539–1555.
27.
BhattacharjeeIBhattacharjaP. Stock price prediction: a comparative study between traditional statistical approach and machine learning approach. In: 2019 4th international conference on electrical information and communication technology (EICT), 2019, pp. 1–6. IEEE.
28.
VijhMChandolaDTikkiwalVA, et al.Stock closing price prediction using machine learning techniques. Procedia Comput Sci2020; 167: 599–606.
29.
MehtabSSenJ, et al. Stock price prediction using machine learning and LSTM-based deep learning models. In: Machine learning and metaheuristics algorithms, and applications: Second symposium, 2021, pp. 88–106.
30.
AssafO. Multivariate and multi-task deep learning architectures for improved stock market prediction and risk management. PhD dissertation, University of Reading, 2024.
31.
TangPTangCWangK. Stock movement prediction: a multi-input LSTM approach. J Forecasting2024; 43: 1199–1211.
32.
DeepA. A multifactor analysis model for stock market prediction. Int J Comput Sci Telecommun2023; 14: 1–6.
33.
AbrahamANathB, et al.Hybrid intelligent systems for stock market analysis. In: Computational science-ICCS: International conference, San Francisco, CA, USA, pp. 337–345, 2001.
34.
ShahDIsahHZulkernineF. Stock market analysis: a review and taxonomy of prediction techniques. Int J Financl Stud2019; 7: 26.
35.
Shui-LingYLiZ. Stock price prediction based on ARIMA-RNN combined model. DEStech Trans Soc Sci Educ Human Sci. 2018: 1–6.
36.
ZhangZ. Research on stock index prediction based on ARIMA-CNN-LSTM model. In: 9th International conference on financial innovation and economic development, pp. 558–565, 2024. Atlantis Press.
37.
TianLFengLYangL, et al.Stock price prediction based on LSTM and LightGBM hybrid model. J Supercomput2022; 78: 11768–11793.
38.
LvPShuYXuJ, et al.Modal decomposition-based hybrid model for stock index prediction. Expert Syst Appl2022; 202: 117252.
39.
LiuZWangY, et al.Kan: kolmogorov-arnold networks. arXiv Preprint2024: arXiv:2404.19756.
40.
Vaca-RubioCJBlancoL, et al.Kolmogorov-arnold networks (kans) for time series analysis. arXiv Preprint2024: arXiv:2405.08790.
41.
SeoKYangJ. Exploring the efficient market hypothesis for accurate stock movement prediction via feature-axis transformer. In Proceedings of the 39th ACM/SIGAPP symposium on applied computing, pp. 892–901, 2024
42.
MaCWuJ, et al.Predicting stock prices using hybrid lstm and arima model.IAENG Int J Appl Math2024; 54: 424–432.
43.
GaoZKuruŏgluEE. Attention based hybrid parametric and neural network models for non-stationary time series prediction. Expert Systems2024; 41: e13419.
44.
HemajothiSBabuSS, et al. Prediction of stock price model through the implementation of hybrid BiLSTM-GRU. In: International conference on recent advances in electrical, electronics, ubiquitous communication, and computational intelligence (RAEEUCCI), Chennai, 2024, pp. 1–6.
45.
WeiYGuXFengZ, et al.Feature extraction and model optimization of deep learning in stock market prediction. J Comput Technol Software2024; 3. https://doi.org/10.5281/zenodo.13622489
46.
Al-KhasawnehMARazaAKhanSUR, et al.Stock market trend prediction using deep learning approach. Comput Econ2024; 64: 1–32.
47.
OyewoleATAdeoyeOBAddyWA, et al.Predicting stock market movements using neural networks: a review and application study. Comput Sci IT Res J2024; 5: 651–670.
48.
ShabanWMAshrafESlamaAE. SMP-DL: a novel stock market prediction approach based on deep learning for effective trend forecasting. Neural Comput Appl2024; 36: 1849–1873.
