A Systematic and Explainable Machine Learning Approach
| *University of Edinburgh Business School | MSc in Data and Decision Analytics (2024/25)* Word count: 16,225 — |
A primary manifestation of long-term climate change is the increased intensity, frequency, and duration of extreme weather events, which already poses a significant negative impact on global agriculture and leads to a long-term threat to worldwide food security. Furthermore, these extreme weather events do not only impact the physical world; they also affect economic and financial markets through a complex transmission pathway.
To address the disconnect between physical climate studies and financial market analyses, this research breaks through the integration gap by building a systematic quantitative framework. By adopting an advanced XGBoost model combined with SHAP values, this study opens the machine learning “black-box” to quantify global and local feature contributions, investigating the complex non-linear interactions of compound weather events, and bridging the yield response signal in the physical world with the pricing mechanism of the financial market.
Note: Click on the chapter links above to navigate through the full manuscript.