A Survey of Explainable Artificial Intelligence (XAI) in Financial Time Series Forecasting

Pierre-Daniel Arsenault,Shengrui Wang,Jean-Marc Patenande

2024-07-23

Abstract:Artificial Intelligence (AI) models have reached a very significant level of accuracy. While their superior performance offers considerable benefits, their inherent complexity often decreases human trust, which slows their application in high-risk decision-making domains, such as finance. The field of eXplainable AI (XAI) seeks to bridge this gap, aiming to make AI models more understandable. This survey, focusing on published work from the past five years, categorizes XAI approaches that predict financial time series. In this paper, explainability and interpretability are distinguished, emphasizing the need to treat these concepts separately as they are not applied the same way in practice. Through clear definitions, a rigorous taxonomy of XAI approaches, a complementary characterization, and examples of XAI's application in the finance industry, this paper provides a comprehensive view of XAI's current role in finance. It can also serve as a guide for selecting the most appropriate XAI approach for future applications.

Machine Learning,Artificial Intelligence

What problem does this paper attempt to address?

The main aim of this paper is to address the issue of the lack of interpretability in artificial intelligence (AI) models for financial time series forecasting. Although AI models have made significant progress in terms of accuracy, their complexity often reduces human trust, especially in high-risk decision-making areas such as finance. To tackle this problem, the paper focuses on the research of explainable artificial intelligence (XAI), particularly XAI methods applied to financial time series forecasting in the past 5 years.

The specific objectives of the paper include:
1. **Provide Definitions**: Distinguish between the concepts of "interpretability" and "explainability" and provide clear definitions for these terms.
2. **Construct a Classification System**: Classify XAI methods based on different principles, such as feature importance, visual explanations, and time series analysis.
3. **Review Existing Technologies**: Review existing interpretable models and explanation methods, including linear regression, attention mechanisms, decision trees, etc., and discuss how these technologies can be used to predict financial time series.
4. **Application Examples**: Showcase practical application cases of XAI in the financial industry to help readers understand how these methods are adopted in practice.
5. **Guidance for Selection**: Provide guidance for data scientists and financial professionals to help them choose the XAI methods that best suit their needs.

Through these efforts, the paper aims to promote responsible and transparent application of AI in the financial sector while enhancing the interpretability and user trust of these systems.