OD41

27 Oktober 2025

Training Machine Learning Models in Power BI

 

Question:

You would like to refresh an ML model to be trained on a new corpus. In which way can you achieve this?

  • Retrain the old model with the new data


Question

:

You are creating an ML model that predicts a categorical value that can belong to one of two categories. Which kind of model is this?

  • Binary classification


Question

:

This is the confusion matrix for a breast cancer prediction model. What is the number of cases where the patient had cancer, but the model predicted that she didn’t?

  • 0


Question

:

This bar chart visualizes the average price of different kinds of cuts for diamonds. Based on this, which type of diamond is most expensive on average?

  • Premium


Question

:

You have created a regression model and would like to evaluate its performance. What is a metric you could use for this?

  • R-squared


Question

:

When working with the Power BI Service, your Power Query processes for loading and transforming data are stored separately in your workspace. What are these stored transformation processes known as?

  • Dataflow


Question

:

This chart visualizes the proportion of correct and incorrect classifications of diamond cuts. Based on this, which one had the highest proportion of incorrect classifications?


  • Very Good


Question

:

Power BI allows you to perform cost-benefit analysis to calculate how profitable a marketing campaign is. This analysis uses a number of different parameters. Please match the input parameter to the statement that best describes it.

  • Population - 

    The number of customers
  • Unit benefit - 

    The reward if a customer responds positively
  • Unit cost - The cost to target each customer


Question

:

This box plot visualizes the difference in insurance premiums for smokers and non-smokers. Based on this, what is the medium premium for smokers?


  • ~35K


Question

:

You are creating an ML model that estimates a continuous value. Which kind of model is this?

  • Regression