Which analytics type involves machine learning on historical data to make predictions?

Predictive analytics involves the use of machine learning techniques applied to historical data to forecast future outcomes or trends. This type of analytics relies on algorithms and statistical models that analyze patterns in existing data to identify the likelihood of various scenarios occurring. By leveraging historical data, predictive analytics can provide insights into what is likely to happen in the future, enabling organizations to make informed decisions based on anticipated trends.

In contrast, descriptive analytics focuses on summarizing historical data to understand what has occurred in the past without attempting to predict future events. Diagnostic analytics goes a step further by examining the reasons behind past outcomes, providing insights into why something happened. Prescriptive analytics, on the other hand, combines data analysis with optimization and simulation techniques to recommend actions based on predicted outcomes but does not specifically focus on making predictions itself.

Given these distinctions, predictive analytics is the correct choice for understanding the application of machine learning on historical data to generate forecasts about future events.