Research on water quality data classification based on Weighted Naive Bayes

Abstract

In order to better implement the water environment management policy , Water quality evaluation is the basic link , That is, according to multiple water quality parameters of a certain water area , How to reasonably divide it into specific water quality categories . For this problem , An improved naive Bayesian classification method is proposed , This method assigns different weights to different attributes , Weaken the assumption of naive Bayesian conditional independence , Make the classification result closer to the actual category . First , Refer to the national surface water quality automatic monitoring station （ Hereinafter referred to as state controlled water station ） Published data , Select one 500 Water quality data as samples , Based on dissolved oxygen 、 Permanganate index 、 Ammonia nitrogen and total phosphorus 4 Establish an evaluation system based on three indicators ; then , Use the improved naive Bayesian classification method to learn and evaluate the samples , And the classification performance is verified by the five fold cross validation method . It turns out that , Improve the accuracy of naive Bayesian classification 、 Accuracy 、 Recall rate and F1 The values reach... Respectively 96.0%、95.9%