Adaptive Machine Learning for Dynamic Financial Markets

Key Points

• The project aims to explore how machine learning models adapt to changes in financial markets over time.
• Utilized structured financial datasets to evaluate model performance.
• Assessed accuracy, stability, and bias of models over time.
• Compared performance between supervised and unsupervised learning paradigms.
• Different machine learning models exhibit varying levels of adaptation to concept drift.
• Some models showed consistent financial decision-making, while others diverged due to constraints.

Abstract