Which of the following four modeling algorithms is least vulnerable to outlier bias? (Select one)A.Linear RegressionB.Naive BayesC.k-NND.GLM

To determine which modeling algorithm is least vulnerable to outlier bias, let's analyze each option:

1. Linear Regression: Linear regression can be significantly affected by outliers, as they can skew the line of best fit. This sensitivity arises because linear regression minimizes the sum of squared errors.

2. Naive Bayes: This algorithm is based on probability and independence assumptions. Naive Bayes is generally less affected by outliers because it relies on the overall distribution of features rather than fitting a specific line or plane.

3. k-NN (k-Nearest Neighbors): k-NN can be affected by outliers, particularly when the outlier is closer to the query point than the majority of true neighbors. This can skew the prediction based on outlier distance.

4. GLM (Generalized Linear Models): Similar to linear regression, GLMs can also be influenced by outliers, depending on the link function used. The extent of the impact can vary based on the distribution of the response variable but is still somewhat sensitive to outliers.

Conclusion

From the analysis, Naive Bayes is the most robust to outliers compared to the other algorithms listed. Therefore, the least vulnerable algorithm to outlier bias is:

Final Answer: B. Naive Bayes