Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Cloud Digital Leader practice bank has an additional 3,070 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A financial services company wants to use machine learning on historical labeled data for two projects:
Which option correctly matches each project to the type of machine learning problem it represents?
Regression is a type of machine learning model that predicts a continuous numerical value, such as how much a customer will spend next month. It is not used to assign records to categories.
Project 1 asked the model to estimate the dollar amount of loss expected from each transaction. A numeric output like that is a regression problem.
A useful test is to ask what the answer looks like. If it is a number on a continuous scale, think regression. If it is a label from a fixed set, think classification, and then count the labels to decide between binary and multi-class.
This option treats every classification task as binary just because each record receives one label. Assigning a single label is true of both binary and multi-class classification, so it does not separate the two types.
the sentiment project only asked whether each comment was positive or not positive. With only two outcomes, both projects would be binary.
Some teams collapse multi-class problems into several binary ones, for example by running separate yes-or-no models for each category. That is a design choice, but it does not change how the problem as stated is classified. Three labels means multi-class.
This option swaps the two problem types. Multi-class classification predicts one class from three or more categories. Binary classification predicts one of exactly two classes.
the labels were reversed. For example, if the fraud team had to sort transactions into several fraud types (stolen card, account takeover, or identity theft), that would be multi-class. If the feedback team only had to decide whether a comment was positive or not, that would be binary.
A common mistake is assuming that anything with "categories" or "detection" must be multi-class. What matters is the count of classes in the problem as defined.
Classification is a type of machine learning that assigns data to discrete categories. It comes in two common forms. Binary classification chooses between exactly two classes, and multi-class classification chooses one class from three or more.
A quick test is to count the possible labels. Two labels means binary. Three or more means multi-class. The fraud scenario has two labels and the sentiment scenario has three. The other options misread either the number of categories or the kind of output the model produces.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated