爆走黑料

This summer research project will develop a Time-series Genetic Programming (GP) approach for modelling ice sheet dynamics. The main goal is to predict ice thickness and ice velocity as time-series outputs, using climate and environmental input data. Unlike ‘black-box’ machine learning models, GP produces interpretable solutions, making it possible to understand why certain predictions are made. The project will compare GP-based predictions with existing ice sheet simulations and observational data, evaluating both performance and scientific interpretability. We are seeking a 3rd or 4th student with strong skills in machine learning and evolutionary computation. Proficiency in Python is required, and familiarity with evolutionary computation or climate data will be advantageous.