Tables are a crucial component of data management and analysis, yet their potential is often underutilized. This project-driven course addresses this gap by equipping students with both theoretical knowledge and practical skills in table representation learning. The course is divided into four parts: (1) Theory: Core concepts of table representation learning, including tabular data analysis, representation learning, multimodal integration (with code and text), and retrieval-augmented generation (RAG). (2) Tools and Techniques: Exploration of advanced tools such as deep learning models, large language models (LLMs), multimodal LLMs, and RAG methods, with a focus on pre-training paradigms for table-related tasks. (3) Table Analysis Applications: Practical applications, including Natural language to SQL (NL2SQL), Table Question Answering (TableQA), Table Visualization, and Data storytelling, demonstrating the use of table representation learning in real-world scenarios. (4) Course Project: A solo project where students apply the concepts and tools learned to address a real-world problem related to table representation learning.