3D data, which is generated by scientific simulations, created by designers, and captured by the real world, is widely used in computer graphics and data visualization. These applications include rendering, 3D object manipulation, augmented reality, and human-AI interaction. With the increasing demand for processing and analyzing such 3D data, there has been significant progress in developing novel technologies, particularly those based on deep learning. This course will explore recent advances in machine learning techniques for 3D data and discuss the remaining challenges. The course will cover basic and advanced machine learning techniques, such as representation learning, neural radiance field, and diffusion model, for 3D data management, rendering, analysis, and generation. The course will be project-oriented, with no exams, but will include manageable programming tasks. It will combine pedagogical lectures, seminar-style reading group presentations, and interactive discussions.