The development of a deep learning based rainfall prediction system
Project Description

Nowadays, many powerful machine learning tools are introduced and are proven to be useful in many fields. However, due to the background knowledge gap between meteorology and computer science, meteorology field still heavily rely on simple models designed by meteorologists based on their experience and observation. Rainfall prediction is proved to be effective with deep learning models. In this project we are focusing on building a user-friendly interface for non-computer science specialists to use such models as well as trying to improve the performance of the current models.

Supervisor
NG Wilfred Siu Hung
Co-Supervisor
NG Wilfred Siu Hung
Quota
2
Course type
UROP1000
UROP1100
Applicant's Roles

In this project you will be closely working with PhD students with strong research and programming skills. You will have the chance to get hands on experience with using well-known popular machine learning (deep learning) tools on a powerful GPU server, and distributed implementation of varied machine learning and data mining tools on a powerful CPU cluster with large datasets. As this project is of both research and application values, it could help in your pursuit of a higher degree or a career in the industry.

We need good CSE background (experience with website construction (in particular Javascript), good programming techniques and basic understanding about machine learning algorithms). You are willing to learn the knowledge and techniques fast in summer if you just basic get some basic understanding of programming and machine learning.

Applicant's Learning Objectives

Develop an interesting interface for the rainfall prediction system with advanced machine learning techniques.

Complexity of the project
Moderate