AI-900 Practice Question: Describe fundamental principles of machine learning on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  A model that predicts which category an item belongs to from a set of options

  Why it's wrong here

  Associating an item with one of several predefined categories is classification, a supervised learning task that outputs discrete class labels, such as 'spam' or 'not spam.' Regression differs fundamentally because it predicts a continuous numerical value rather than a discrete label, so the two tasks require different evaluation metrics (e.g., accuracy versus root mean squared error). While logistic regression has 'regression' in its name, it is actually a classification algorithm.

• ✓

  Predicting a continuous numerical value such as price, temperature, or demand

  Why this is correct

  Regression in machine learning is the task of predicting a continuous numerical value, such as house price, energy demand, or tomorrow's temperature, from input features. The model learns a mapping from features to a real-valued output and is typically trained with loss functions like mean squared error to penalize the magnitude of prediction errors. This contrasts with classification, which predicts discrete categories, and is used in scenarios where the outcome naturally exists on an infinite numeric scale.

• ✗

  Going back to a previous model version when the current version performs poorly

  Why it's wrong here

  Rolling back to a previous model version when the current one underperforms is an MLOps deployment process, not a machine learning task that defines regression. Regression is a supervised learning paradigm where the target label is a continuous numeric quantity, such as price, temperature, or demand, and the model is trained with a loss function that measures distance between predicted and actual values. Reverting a deployed endpoint is a version-control and operational action, unrelated to modeling type.

• ✗

  A technique for reducing the dimensionality of training data before model fitting

  Why it's wrong here

  Reducing dimensionality before model fitting, using techniques like principal component analysis (PCA) or autoencoders, is an unsupervised feature-engineering step aimed at compressing the input space while retaining variance. Regression, in contrast, is a supervised learning task that predicts a continuous numerical outcome from input features; it does not transform data or reduce feature counts. Dimensionality reduction often precedes any predictive modeling, including regression and classification, as a data preprocessing step.

Go deeper

Related to this question