Which aspect does high variance in a model negatively impact?

• A. Model consistency across different datasets
• B. Model interpretability
• C. Model training time
• D. Model complexity

Answer: (A)

High variance in a model negatively impacts model consistency across different datasets. When a model has high variance, it becomes overly sensitive to the fluctuations in the training data, which leads to a scenario known as overfitting. This means that while the model performs exceptionally well on the training dataset, its performance diminishes significantly when it encounters new, unseen data. As a result, the predictions made by the model may vary widely depending on the specific dataset being used, thus compromising the model's reliability and generalization abilities.

This inconsistency across different datasets reflects a lack of robustness, as the model fails to capture the underlying relationships in the data, leading to biased or inaccurate predictions when applied to real-world situations.