数据挖掘：C5.0决策树算法在警察院校学生体质分析中的应用

Research of Data-mining in Police Training: The Application of C5.0 Decision Tree to Students ’ Constitution in Police Colleges

C5.0决策树算法适用于大数据集处理,特别是它的Boosting集成机器学习算法可以有效地将精度较低的"弱学习算法"提升为精度较高的"强学习算法",从而达到模型修剪与优化的目的。研究结果表明:C5.0决策树算法生成的模型可以精确地评价学生的体质健康状况(97.8%)且模型预测的泛化能力较强(98.1%)。因此,C5.0决策树算法可以用来判断影响警察院校学生体质测试成绩的关键因素,为深层挖掘相关警务数据内涵与监测提供了实证依据。

The C5.0 Decision Tree can be used for large data sets. Due to the addition of Boosting, The C5.0 Decision Tree can get better models. Booting optimizes the model by effectively improving the less accurate "weak learning algorithm" to a more accurate "strong learning algorithm."Result: The decision tree model generated by C5.0 algorithm can accurately evaluate students’ physical health status(98.4%) with simple physical monitoring indicators and the generalization of model prediction can be strong(98.2%). Conclusion: The C5.0 algorithm can be used to determine the key factors physical test results, to deeply dig Students ’ Constitution data and monitor its changes in Police Colleges.