Application of Feature Weighted KNN Classification Algorithm in Professional Curriculum Talent Training

In order to promote professional teaching reform in universities and improve the quality of talent cultivation in professional courses, information and data of college freshmen are collected and maintained. Through the enrollment system, specialized courses suitable for students are selected, which are a typical application scenario for big data management and analysis. The purpose of this article was to study the application of feature weighted KNN (K-Nearest Neighbor) classification algorithm in professional curriculum talent cultivation. A professional course recommendation system based on feature weighted KNN classification algorithm was designed. The system first determined the attributes of the recommended samples from aspects such as relevance to student samples. Then, on this basis, computing tools such as HDFS (Hadoop Distributed File System), HBase, and MapReduce were integrated, and the feature weight KNN algorithm customized by MapReduce was utilized to process samples in a distributed manner to form an accurate recommendation model. According to the results of algorithm comparison, the classification accuracy of the feature weighted KNN classification algorithm designed in this paper was above 80%.