Course:VANT149/2024/Capstone/Science/Team10

Abstract

The study investigates the efficacy of various probabilistic models in predicting precipitation by analyzing a decade's weather data from Vancouver. Through rigorous comparison of Markov, Hidden Markov, and Bayesian models, the research highlights the strengths and limitations of each approach. Initial results show the Bayesian model, supplemented with historical data insights, significantly outperforms other models in forecasting accuracy. This research not only for advances precipitation forecasting but also designed as an educational tool, enhancing understanding of statistical models and their applications in meteorology.

Key words: Precipitation prediction, Markov model, Hidden Markov model (HMM), Bayesian model.

Biographies

Tianchen Liao

Tianchen Liao is a first-year science student at the University of British Columbia in the Vantage One Program. His research interests include mathematics and computer science, particularly in data science and cognitive science. Currently, he is working on a project to evaluate the ability of various probabilistic models to predict rainfall, and he also leads a music studio in his spare time.

Zihan Zhou

Zihan Zhou is a student interested in computer science and statistics in the Vantage One Science Program at the University of British Columbia. She aims to evaluate the ability of precipitation prediction of various models as the solution for the influence of significant changes in regional rainfall on human activities in recent decades. Her main contributions to the project include filtering original data with computer science knowledge and sorting and labeling data with statistical knowledge.