Higher vocational education is a self-consistent system for higher education appropriate to the development of productivity and economy in the world. It aims at training skilled talents, which has made great contribution to the economy and industry. Generally, designing courses in high vocational education includes teaching analysis, teaching strategy, teaching practice and teaching assessment. Among the teaching steps, teaching assessment is one of the most important method to improve the quality of course teaching. However, in most high vocational education courses, traditional written exam is still the primary tools of assessments, which can not fulfill the development of high vocational education. In order to improve the quality of high vocational education, it is very urgent to design a practical and efficient system with multiple assessments. We exploit machine learning techniques to design assessment system for high vocation education. Machine learning is a very powerful tool for data analysis and it has been used for education tools in recent years. First, we improve the teaching organization for training skilled talents. Second, we propose a feature selection model based on the improved teaching organization. Third, we propose a machine learning model for teaching assessment. With the main contributions and other improvements, we design a multi-assessment system for vocational teaching based on machine learning. We implement the multi-assessment system by using Python and TensorFlow, which shows that the system can provide practical and efficient multiple assessments for vocational teaching based on training machine learning model. Compared with other assessment methods, machine learning based multi-assessment is more intelligent and automatic. Besides, it can be extended to other fields of education with slight modifications.

高等职业教育是适合于世界生产力和经济发展的高等教育的自洽体系。它旨在培养技术人才，为经济和工业做出了巨大贡献。一般来说，设计高等职业教育的课程包括教学分析，教学策略，教学实践和教学评估。在教学步骤中，教学评估是提高课程教学质量的最重要方法之一。但是，在大多数高等职业教育课程中，传统的笔试仍然是考核的主要手段，无法满足高等职业教育的发展。为了提高高职教育的质量，迫切需要设计一种实用，高效，有多种评估的系统。我们利用机器学习技术来设计高职教育评估系统。机器学习是用于数据分析的非常强大的工具，并且近年来已用于教育工具。首先，我们完善了培训技能型人才的教学组织。其次，我们提出了一种基于改进的教学组织的特征选择模型。第三，我们提出了一种用于教学评估的机器学习模型。在主要贡献和其他改进的基础上，我们设计了一种基于机器学习的职业教育多评估系统。我们使用Python和TensorFlow实现了多评估系统，表明该系统可以基于训练机器学习模型为职业教学提供实用，高效的多重评估。与其他评估方法相比，基于机器学习的多重评估更加智能和自动化。此外，只要稍加修改，它就可以扩展到其他教育领域。