With the widespread application of big data and artificial intelligence, quantitative analysis methods have become increasingly important in urban and rural planning education. However, current teaching still faces multiple challenges: the statistical knowledge system is abstract and lacks coherence; case studies often fail to reflect the context of urban–rural research; theoretical instruction and software operation are difficult to balance within limited class hours; and after-class exercises lack effective feedback. To address these issues, this paper proposes several strategies, including differentiated learning objectives, discipline-specific case-driven teaching, a division of labor between classroom theoretical instruction and after-class software practice, and feedback-oriented exercises, supported by the development of an intelligent teaching platform. The platform’s resource system integrates lecture slides, datasets, and software operation manuals, thereby creating a coherent link between classroom learning and self-study. Its exercise system employs a mechanism of “one student, one dataset, automatic grading, and real-time feedback”, providing personalized training and instant correction. This not only effectively prevents plagiarism but also significantly enhances students’ concentration and mastery. Teaching practice at Soochow University demonstrates that this model markedly promotes students’ initiative and comprehensive application ability, fostering a virtuous cycle of “understanding–practice–feedback–improvement”. The findings provide a feasible path for reforming the teaching of quantitative analysis methods in urban and rural planning and offer useful insights for the intelligent transformation of other planning-related courses.