Describe fundamental principles of machine learning on Azure →easyMultiple ChoiceObjective-mapped

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

What is a training dataset in machine learning?

⚠ Common exam trap

Watch out — candidates often confuse the training dataset with the test dataset or preprocessed data, mistakenly thinking any cleaned data or evaluation data qualifies as training data, when in fact the training dataset is specifically the labeled subset used to fit the model's parameters.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

The labeled data used to teach a machine learning model

A training dataset is the labeled data used to teach a machine learning model by allowing it to learn patterns and relationships between features and labels. In Azure Machine Learning, this dataset is fed into an algorithm during the training step, where the model adjusts its internal parameters (e.g., weights in a neural network) to minimize prediction error. Without labeled training data, supervised learning models cannot learn the mapping from inputs to outputs.

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 dataset used to evaluate a trained model's performance on unseen data
Why it's wrong here
This describes the test dataset, which is held out from the training process and used only once to measure how well the model generalizes to new, unseen examples. Unlike training data, test data never participates in parameter updates or loss minimization. In supervised learning, training data provides the labeled examples the model actively learns from, so the model cannot be taught by a dataset that is deliberately kept out of the fitting process.
✓
The labeled data used to teach a machine learning model
Why this is correct
Training data consists of input features paired with known correct output labels, serving as the ground truth for supervised learning. During training, the model iteratively adjusts its internal parameters (such as weights in a neural network) to minimize prediction error on these examples, thereby learning the underlying patterns and relationships. This labeled data is the direct source of knowledge for the model, distinguishing it from unlabeled inference data or preprocessed data that has not been used for teaching.
✗
Data that has been cleaned and normalized for analysis
Why it's wrong here
Cleaned and normalized data is the result of preprocessing steps such as handling missing values, scaling features, or removing outliers, which are applied to raw data before modeling. However, the training dataset is a specific subset of preprocessed examples reserved for teaching the model, not merely any tidy dataset. Preprocessing applies to both training and test data uniformly, but only the labeled training subset is used to update the model's weights, so this definition confuses data preparation with the purpose of the training set.
✗
Real-world data used after model deployment
Why it's wrong here
This refers to production or inference data, which is real-world, typically unlabeled input encountered after the model is deployed to make predictions. This data is used only for generating outputs, never for updating the model's learned parameters, because the training phase has already ended. Training data, in contrast, contains labels and is consumed during the learning phase to shape the model's decision boundaries, so it is fundamentally different from post-deployment inference streams.

Go deeper

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

Why each option matters

Option-by-option breakdown

Related to this question