A/B testing compares multiple versions of a webpage to measure performance differences. 

Logistic regression is suitable for predicting continuous outcomes.  

Social Network Analysis (SNA) considers relationships and connections between users.

Which evaluation metric is most appropriate for imbalanced classification problems?

One-Hot Encoding is typically used for:

Principal Component Analysis (PCA) reduces dimensionality by identifying directions of maximum variance. 

Time-series anomaly detection often uses:

Tokenization splits text into smaller units such as words or subwords.

Machine-generated data is often real-time and massive in scale.

A confusion matrix with many false negatives suggests: