KNN算法求解

最佳参数K已经确定
Model_KNN = ClassificationKNN.fit(xtrain,ytrain,'NumNeighbors',Kbest, 'BreakTies','nearest','Distance',DistanceMethod);
PredictTestLabels_KNN = predict(Model_KNN, xtest);
NegAcc = sum( PredictTestLabels_KNN== 0)/sum( ytest==0);
PosAcc = sum( PredictTestLabels_KNN== 1)/sum( ytest==1);
F1score = NegAcc*PosAcc/(NegAcc+PosAcc)
这里的NegAcc是负样本的准确率吗，怎么通过这个准确率推导TN TP FN FP四个指标呢。