Research on Medical Information Classification and Extraction Based on Optimized KNN Algorithm

The rapid advancement of big data technology poses challenges to medical information management, especially in terms of data diversity and structural complexity. This study focuses on the issue of missing values in medical laboratory test data and designs a missing value imputation scheme mainly based on the KNN algorithm. It uses weighted Euclidean distance to measure sample similarity and combines the features of neighboring samples to construct a dynamic filling model. Particularly, a feature weight adaptive adjustment strategy is introduced in the nearest neighbor selection mechanism, effectively enhancing the modeling effect of complex medical datasets. Seven standard medical datasets from the UCI database were selected for experiments, and the performance metrics of the KNN algorithm, improved weighted KNN model, random forest, support vector machine, and naive Bayes classifier were compared horizontally ^{[<xref ref-type="bibr" rid="R1">1</xref>]}. By improving feature engineering, the aim is to enhance the quality of medical data and optimize clinical decision support systems. These technological breakthroughs provide key support for the construction of intelligent medical data analysis platforms and are of great significance for the advancement of precision medicine.