在 R 中使用 naivebayes 和 Caret 实现朴素贝叶斯分类器

发布于 2025-01-12 07:33:12 字数 681 浏览 0 评论 0原文

我有一个使用 naivebayes 包的现有预测模型。我目前正在尝试做的是使用 Caret 复制我的模型。这样做的原因是因为我想运行 varImp() 函数来查看插入符号中的重要变量列表。我已经成功地使用 Caret 运行朴素贝叶斯分类器，但问题是，当我进行预测以确保使用 Caret 的模型与使用 naivebayes 的模型一致时，我最终会得到不同的结果。有人可以建议我如何使用 Caret 复制现有的朴素贝叶斯分类器吗？

# Naive Bayes Classifier using naivebayes
nb = naive_bayes(y ~ ., data = df, usekernel = T)

# Naive Bayes Classifier using Caret
Grid = data.frame(usekernel = TRUE, laplace = 0, adjust = 1)
caret = train(y ~ ., data = df,
              method = 'naive_bayes',
              na.action = na.pass,
              trControl = trainControl(method = "none"),
              tuneGrid = Grid)

提前非常感谢。

谢谢！

原文

I have an existing predictive model using naivebayes package. What I'm currently trying to do is replicate my model using Caret. The reason for this is because I'd like to run varImp() function to see the list of significant variables from Caret. I have managed to run Naive Bayes Classifier using Caret, but the problem is that when I do the prediction to make sure the model from using Caret aligns with the model from using naivebayes, I end up with different results. Could someone please advice how I can replicate my existing Naive Bayes Classifier using Caret?

# Naive Bayes Classifier using naivebayes
nb = naive_bayes(y ~ ., data = df, usekernel = T)

# Naive Bayes Classifier using Caret
Grid = data.frame(usekernel = TRUE, laplace = 0, adjust = 1)
caret = train(y ~ ., data = df,
              method = 'naive_bayes',
              na.action = na.pass,
              trControl = trainControl(method = "none"),
              tuneGrid = Grid)

Much appreciated in advance.

Thanks!

分享到QQ

分享到微博