Transfer learning for financial data predictions: a systematic review

ArXiv ID: 2409.17183 “View on arXiv”

Authors: Unknown

Abstract

Literature highlighted that financial time series data pose significant challenges for accurate stock price prediction, because these data are characterized by noise and susceptibility to news; traditional statistical methodologies made assumptions, such as linearity and normality, which are not suitable for the non-linear nature of financial time series; on the other hand, machine learning methodologies are able to capture non linear relationship in the data. To date, neural network is considered the main machine learning tool for the financial prices prediction. Transfer Learning, as a method aimed at transferring knowledge from source tasks to target tasks, can represent a very useful methodological tool for getting better financial prediction capability. Current reviews on the above body of knowledge are mainly focused on neural network architectures, for financial prediction, with very little emphasis on the transfer learning methodology; thus, this paper is aimed at going deeper on this topic by developing a systematic review with respect to application of Transfer Learning for financial market predictions and to challenges/potential future directions of the transfer learning methodologies for stock market predictions.

Keywords: Transfer Learning, systematic review, neural networks, financial time series, stock market predictions, Equities (Stocks)

Complexity vs Empirical Score

• Math Complexity: 4.0/10
• Empirical Rigor: 2.0/10
• Quadrant: Philosophers
• Why: The paper is a systematic review and theoretical discussion of transfer learning methodologies, relying on high-level descriptions of models (e.g., neural networks, GARCH) without presenting new mathematical derivations or complex equations. Empirical rigor is low because it synthesizes existing literature rather than presenting original backtests, datasets, or implementation-heavy experiments.

  flowchart TD
    A["Research Goal<br/>'What are the applications and challenges<br/>of Transfer Learning in stock price prediction?'"] --> B["Methodology<br/>Systematic Literature Review"]
    B --> C["Data Source<br/>Existing Neural Network Models for Finance"]
    C --> D{"Computational Process<br/>Analyze TL Methods & Performance"}
    D --> E{"Challenge?"}
    E -- Yes --> F["Outcome: Limitations<br/>Domain Shift & Data Scarcity"]
    E -- No --> G["Outcome: Key Findings<br/>TL overcomes data scarcity<br/>and improves accuracy vs. traditional ML"]
    F --> H["Future Directions<br/>Deep TL & Domain Adaptation"]
    G --> H
    H --> I["Conclusion<br/>TL is vital for financial prediction<br/>(requires domain-specific design)"]