Master the AI Frontier 2026: Rock Your Machine Learning & NLP Skills with This Epic Practice Test!

Predicting future outcomes from historical data relies on learning patterns from past observations and applying that knowledge to new cases. This process is called prediction, the general term for estimating unknown values or labels for future data points based on a trained model.

A common specialization of prediction is forecasting, which is especially used for time-series data where the goal is to estimate future values over time. Classification, in contrast, assigns data to predefined categories. Clustering groups similar data points without using predefined labels. Inference involves drawing conclusions about the underlying data-generating process or its parameters.