Paper: arXiv 2501.09760
Authors: Yuxi Hong
Abstract
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 R^2 as high as 0.98, when compared with other competitive networks.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 2.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper employs advanced math including Transformer encoders, BiGRU, and Kolmogorov-Arnold Networks with complex formulations, but lacks concrete backtesting details, code, or robust data implementation, focusing instead on reported metrics without addressing real-world deployment challenges.
Research Flowchart
flowchart TD
A["Research Goal: Accurate Stock Market Prediction"] --> B["Input Data: Financial Time-Series"]
B --> C["Method: Multi-Layer Hybrid MTL Structure"]
C --> D["Transformer Encoder: Extract Feature Correlations"]
D --> E["BiGRU: Capture Temporal Dependencies"]
E --> F["KAN: Enhance Non-linear Learning"]
F --> G{"Outcomes & Metrics"}
G --> H["MAE: 1.078"]
G --> I["MAPE: 0.012"]
G --> J["R^2: 0.98"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a novel multi-layer hybrid multi-task learning (MTL) framework that combines a Transformer encoder, BiGRU, and KAN for stock market prediction.
Source sentence
This paper proposes a novel multi-layer hybrid multi-task learning (MTL) framework aimed at achieving more efficient stock market predictions.
Assumes
No explicit assumptions found in the text.
Data
No dataset named in the text.
Costs
Not applicable
Finds 3
- The proposed model achieves an MAE as low as 1.078, a MAPE as low as 0.012, and an R² as high as 0.98 compared with other competitive networks. performance
Source sentence
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.
- In further experiments the method attains MAE 0.45 and R² up to 0.98, highlighting its robustness and predictive accuracy. predictive-accuracy
Source sentence
Experimental results demonstrate that the proposed learning network outperforms competitive models, achieving MAE as low as 0.45 and R2 up to 0.98, highlighting its robustness and predictive accuracy.
- The proposed method reduces RMSE to 39.820, a 17.2% reduction versus a baseline, and achieves R² of 0.977, outperforming prior works by several percentage points. performance
Source sentence
Specifically, the RMSE of the proposed perdition method is 39.820, indicating a 17.2% reduction compared to Hema‑jothi et al. [44]
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 76 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (44)
- A Stock Market Trends Analysis of Reliance using Machine Learning Techniques (2023)
- Progress and prospects of data-driven stock price forecasting research (2023)
- A study of the impact of COVID-19 on the Chinese stock market based on a new textual multiple ARMA model (2023)
- Stock price prediction using the ARIMA model (2014)
- Stock return prediction under GARCH—An empirical assessment (2017)
- Forecasting Indonesian Stock Index Using ARMA-GARCH Model (2024)
- Comparison of ARIMA, ANN and LSTM for stock price prediction (2020)
- China’s commercial bank stock price prediction using a novel K-means-LSTM hybrid approach (2022)
and 36 more
- A new CNN-based model for financial time series: TAIEX and FTSE stocks forecasting (2022)
- Stock Price Prediction Using RNNs: A Comparative Analysis of RNN, LSTM, GRU, and BiRNN (2024)
- A graph-based CNN-LSTM stock price prediction algorithm with leading indicators (2023)
- Graph-based stock correlation and prediction for high-frequency trading systems (2022)
- Integrated GCN-LSTM stock prices movement prediction based on knowledge-incorporated graphs construction (2024)
- Stock market index prediction using transformer neural network models and frequency decomposition (2024)
- Data-driven stock forecasting models based on neural networks: A review (2024)
- Achieving more with less: A tensor-optimization-powered ensemble method (2024) — link
- A Margin-Maximizing Fine-Grained Ensemble Method (2024) — link
- A survey on ensemble learning (2020)
- Application of an ANN and LSTM-based Ensemble Model for Stock Market Prediction (2024)
- Stock market prediction with time series data and news headlines: a stacking ensemble approach (2024)
- An effective time series analysis for stock trend prediction using ARIMA model for nifty midcap-50 (2013)
- Stock price prediction using ARIMA model (2021)
- Prediction of Popular Global Stock Indexes Volatility by Using ARCH/GARCH Models (2024)
- Stock market prediction using hidden markov models (2012)
- Research on HMM-Based Efficient Stock Price Prediction (2022)
- Network log-ARCH models for forecasting stock market volatility — → 2303.11064
- Stock price prediction: a comparative study between traditional statistical approach and machine learning approach (2019)
- Stock closing price prediction using machine learning techniques (2020)
- Stock price prediction using machine learning and LSTM-based deep learning models (2021)
- Multivariate and multi-task deep learning architectures for improved stock market prediction and risk management (2024)
- Stock movement prediction: A multi-input LSTM approach (2024)
- A Multifactor Analysis Model for Stock Market Prediction (2023)
- Hybrid intelligent systems for stock market analysis (2001)
- Stock market analysis: A review and taxonomy of prediction techniques (2019)
- Stock price prediction based on ARIMA-RNN combined model (2017)
- Research on stock index prediction based on ARIMA-CNN-LSTM model (2024)
- Stock price prediction based on LSTM and LightGBM hybrid model (2022)
- Modal decomposition-based hybrid model for stock index prediction (2022)
- Kan: Kolmogorov-arnold networks (2024) — link
- Kolmogorov-arnold networks (kans) for time series analysis (2024) — link
- Exploring The Efficient Market Hypothesis for Accurate Stock Movement Prediction via Feature-Axis Transformer (2024)
- Predicting Stock Prices Using Hybrid LSTM and ARIMA Model. (2024)
- Attention based hybrid parametric and neural network models for non-stationary time series prediction (2024)
- Prediction of Stock Price Model through the Implementation of Hybrid BiLSTM-GRU (2024)
1 of 44 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.