Courseiva

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

What is 'stochastic gradient descent' (SGD) and how does it work?

⚠ Common exam trap

Many candidates confuse 'stochastic' with 'random sampling of data' (Option A) or 'random model selection' (Option C), when in fact SGD's stochasticity refers to using random mini-batches to compute gradients, not random data selection or architecture choice.

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

An optimisation algorithm that updates weights using gradients computed on random data mini-batches

Stochastic Gradient Descent (SGD) is an optimization algorithm used to train machine learning models by iteratively updating model weights. It computes the gradient of the loss function on a randomly selected mini-batch of training data (not the entire dataset), which introduces noise but significantly speeds up convergence and reduces memory usage. This mini-batch approach is the core of SGD and distinguishes it from batch gradient descent.

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 random sampling method for selecting training data without replacement

    Why it's wrong here

    This describes a data-subsetting strategy, such as without-replacement sampling used when partitioning a dataset into training and validation folds. SGD selects random mini-batches only to estimate the loss gradient cheaply for a single weight update; the mini-batches are not permanently withheld from training and the procedure is not a general method for dataset selection.

  • An optimisation algorithm that updates weights using gradients computed on random data mini-batches

    Why this is correct

    Stochastic Gradient Descent (SGD) is an iterative optimization algorithm where each update step uses a randomly drawn mini-batch of training examples to compute an unbiased estimate of the loss gradient. Because the mini-batch is small and randomly chosen, the gradient estimate is noisy, but it is far cheaper to compute than the full-batch gradient, allowing SGD to scale to large datasets and still converge by taking many noisy steps.

  • A technique for randomly selecting which model architecture to use for AutoML

    Why it's wrong here

    This describes model search in AutoML, where different neural network architectures (such as ResNet versus EfficientNet) are evaluated for a task. Stochastic Gradient Descent (SGD) is not an architecture selection technique; it is an optimizer that adjusts the weights of an already chosen architecture by taking steps in the direction of the negative gradient of the loss function.

  • Randomly descending through decision tree branches to make predictions

    Why it's wrong here

    Traversing a trained decision tree from root to leaf by applying feature-value comparisons at each branch is how predictions are made at inference time. SGD, by contrast, is a weight-update procedure that iteratively minimizes a loss function using gradient information, and it has no relation to how tree-based models make predictions.

About these practice questions

This AI-900 question is part of Courseiva's 985-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

JA

Written by Johnson Ajibi, MSc IT Security

Senior Network & Security Engineer · founder of Courseiva

This AI-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-900 exam.