Which of the following is a supported model type in PMML?

In the context of Predictive Model Markup Language (PMML), the supported model types include a variety of analytical models used in data science and analytics. The scorecard model is specifically designed for representing scoring mechanisms typically used in credit risk applications or other binary classification tasks. This model type is aligned with the capabilities of PMML, which aims to standardize the way to represent various predictive models.

Scorecards can provide a weighted score based on the characteristics or features input into the model, allowing organizations to quantitatively assess risks or probabilities. Thus, it fits well within the structure and intentions of PMML as it provides a clear methodology for calculating scores based on inputs.

In contrast, while data mining models, entity resolution models, and time series models are certainly significant in the field of data analytics, they may not be as explicitly defined or supported in PMML as scorecards are. For instance, although data mining might encompass a variety of techniques, PMML focuses on specific models rather than the broader category. Entity resolution and time series models may also exist but do not reflect the core elements PMML was designed to represent, which is why the scorecard stands out as the accurate answer.