Supervised machine learning

What is variance in the context of machine learning?

• Variance is a measure of a model's sensitivity to changes in the training dataset, indicating overfitting or generalization.

• Variance is the total number of features in a dataset.

• Variance measures the average error of a model's predictions.

• Variance indicates the speed of a model's training process.

How does high variance affect a model's performance?

• High variance improves a model's performance by enhancing accuracy on training data.

• High variance negatively impacts a model's performance by causing overfitting, leading to poor generalization on unseen data.

• High variance leads to underfitting, resulting in better generalization on unseen data.

• High variance has no effect on a model's performance regardless of the data.

What is the relationship between bias and variance?

• Bias and variance are two sources of error in machine learning models, where bias leads to underfitting and variance leads to overfitting.

• Variance is always beneficial for model performance.

• Bias is the only source of error in models.

• Bias and variance are unrelated concepts.

Explain the concept of overfitting in machine learning.

• Overfitting is when a model learns the training data too well, leading to poor performance on new data.

• Overfitting happens when a model is trained on too little data, leading to generalization issues.

• Overfitting is when a model performs equally well on both training and new data.

• Overfitting occurs when a model is too simple and cannot capture the underlying patterns.

What are some common signs of overfitting in a model?

• Common signs of overfitting include high training accuracy with low validation accuracy, poor performance on unseen data, and overly complex model predictions.

• High validation accuracy with low training accuracy

• Simple model predictions with no variance

• Consistent performance on unseen data

How can you reduce variance in a machine learning model?

• Use regularization, simplify the model, increase training data, or use ensemble methods.

• Ignore feature selection

• Use fewer training examples

• Increase model complexity

What role does regularization play in controlling variance?

• Regularization only affects bias and not variance.

• Regularization increases variance by allowing more complex models.

• Regularization has no effect on model performance or variance.

• Regularization reduces variance by penalizing model complexity, preventing overfitting.

Define the term 'generalization error' in machine learning.

• The generalization error is the difference between training and validation accuracy.

• The generalization error is the measure of a model's ability to perform well on unseen data compared to its performance on the training data.

• The generalization error is the measure of a model's complexity and size.

• Generalization error refers to the amount of data used for training a model.

What is the purpose of cross-validation in assessing model performance?

• To ensure the model only performs well on the training data.

• To evaluate model performance and ensure it generalizes well to unseen data.

• To reduce the amount of data used for training.

• To increase the training time of the model.

How does the Occam's Razor principle relate to variance?

• Occam's Razor promotes the use of multiple competing theories to explain variance.

• Occam's Razor is unrelated to model simplicity and variance.

• Occam's Razor suggests using complex models to reduce variance.

• Occam's Razor advocates for simpler models to explain variance effectively.