Find a practice exam
Flexible search understands AI-901, ai901, ai 901, 901, ai, network plus, and saa c03.
No matching practice exams yet.
Stanford Machine Learning Practice Test
Start a free 30-question Stanford Machine Learning daily set with source-backed explanations, local progress, and a fresh rotation every morning.
Stanford Machine Learning
Stanford / DeepLearning.AI Machine Learning Specialization
Why this page works
- Thirty focused questions every day
- Source links on every explanation
- Local progress saved automatically
- Email sync path ready for later
- Apps provide deeper drills when available
Use this Stanford Machine Learning practice test to review Machine Learning Specialization. Questions rotate daily and each explanation links to the source used to validate the answer.
7-day score keeper
Answer questions today and this will become a rolling 7-day scorecard.
Keep today’s practice moving
Guest progress saves automatically on this device. Add an email later when you want a magic link that keeps your daily ML Specialization practice in sync across browsers.
Guest progress saves on this device automatically
30 verified questions are currently in the live bank. Questions updated at Apr 13, 2026, 10:51 AM CDT. The daily set rotates at 10:00 AM local time, and each explanation links back to the source used to write it. Use the web set for quick practice, then switch to the app when available for larger banks and deeper review.
Use these official DeepLearning.AI resources alongside the daily practice set. They cover the provider's own exam page, study guide, or prep material.
Need adjacent DeepLearning.AI practice pages too? DeepLearning.AI practice hub.
Which answer is the best source-backed summary of this Stanford / DeepLearning.AI Machine Learning Specialization topic?
A. Correct: Model evaluation compares performance on data not used to directly fit the model is the correct answer because model evaluation compares performance on data not used to directly fit the model. Practical ML requires measuring how well a model generalizes.
B. Incorrect: Evaluation should only use memorized training examples is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
C. Incorrect: Generalization is unrelated to ML quality is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
D. Incorrect: Metrics are never useful in ML is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
A learner is reviewing Stanford-ML-neural-networks. What should they remember?
A. Incorrect: Layers and parameters never affect outputs is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
B. Correct: Neural networks learn parameters across layers to model complex relationships is the correct answer because neural networks learn parameters across layers to model complex relationships. Neural networks are part of the specialization path.
C. Incorrect: Neural networks are only physical cabling diagrams is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
D. Incorrect: Neural networks cannot be trained from data is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
What is the safest study takeaway for Supervised Learning?
A. Correct: Supervised learning trains models from examples that include input features and target labels is the correct answer because supervised learning trains models from examples that include input features and target labels. Supervised learning is a foundational topic in the specialization.
B. Incorrect: Supervised learning never uses labels is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
C. Incorrect: Supervised learning is only file encryption is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
D. Incorrect: Input features are unrelated to supervised models is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
Which statement best matches Unsupervised Learning for Stanford Machine Learning practice?
A. Incorrect: Pattern discovery is unrelated to unsupervised learning is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
B. Incorrect: Unsupervised learning always requires target labels is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
C. Correct: Unsupervised learning finds patterns in data without target labels is the correct answer because unsupervised learning finds patterns in data without target labels. Unsupervised learning is a core ML category.
D. Incorrect: Unsupervised learning is only a keyboard shortcut is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
When practicing Stanford Machine Learning, which option belongs under Practical Advice?
A. Incorrect: Iteration priorities are unrelated to model improvement is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
B. Incorrect: Practical ML advice means never diagnosing errors is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
C. Incorrect: Bias and variance cannot affect models is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
D. Correct: Practical ML work diagnoses bias, variance, data issues, and iteration priorities is the correct answer because practical ML work diagnoses bias, variance, data issues, and iteration priorities. The specialization includes practical ML decision-making concepts.
Which statement best matches Recommender Systems for Stanford Machine Learning practice?
A. Incorrect: Recommender systems only assign IP addresses is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
B. Incorrect: Recommender systems cannot use data is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
C. Incorrect: User preferences are unrelated to recommendations is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
D. Correct: Recommender systems predict user preferences or item relevance from data is the correct answer because recommender systems predict user preferences or item relevance from data. Recommender systems are a common ML application area.
When practicing Stanford Machine Learning, which option belongs under Model Evaluation?
A. Incorrect: Generalization is unrelated to ML quality is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
B. Incorrect: Evaluation should only use memorized training examples is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
C. Incorrect: Metrics are never useful in ML is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
D. Correct: Model evaluation compares performance on data not used to directly fit the model is the correct answer because model evaluation compares performance on data not used to directly fit the model. Practical ML requires measuring how well a model generalizes.
Which statement best matches Neural Networks for Stanford Machine Learning practice?
A. Incorrect: Layers and parameters never affect outputs is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
B. Correct: Neural networks learn parameters across layers to model complex relationships is the correct answer because neural networks learn parameters across layers to model complex relationships. Neural networks are part of the specialization path.
C. Incorrect: Neural networks are only physical cabling diagrams is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
D. Incorrect: Neural networks cannot be trained from data is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
Which statement best matches Supervised Learning for Stanford Machine Learning practice?
A. Incorrect: Supervised learning never uses labels is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
B. Incorrect: Supervised learning is only file encryption is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
C. Correct: Supervised learning trains models from examples that include input features and target labels is the correct answer because supervised learning trains models from examples that include input features and target labels. Supervised learning is a foundational topic in the specialization.
D. Incorrect: Input features are unrelated to supervised models is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
Which answer is the best source-backed summary of this Stanford / DeepLearning.AI Machine Learning Specialization topic?
A. Incorrect: Unsupervised learning always requires target labels is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
B. Incorrect: Pattern discovery is unrelated to unsupervised learning is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
C. Incorrect: Unsupervised learning is only a keyboard shortcut is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
D. Correct: Unsupervised learning finds patterns in data without target labels is the correct answer because unsupervised learning finds patterns in data without target labels. Unsupervised learning is a core ML category.
Which statement best matches Practical Advice for Stanford Machine Learning practice?
A. Incorrect: Iteration priorities are unrelated to model improvement is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
B. Incorrect: Practical ML advice means never diagnosing errors is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
C. Incorrect: Bias and variance cannot affect models is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
D. Correct: Practical ML work diagnoses bias, variance, data issues, and iteration priorities is the correct answer because practical ML work diagnoses bias, variance, data issues, and iteration priorities. The specialization includes practical ML decision-making concepts.
What is the safest study takeaway for Recommender Systems?
A. Correct: Recommender systems predict user preferences or item relevance from data is the correct answer because recommender systems predict user preferences or item relevance from data. Recommender systems are a common ML application area.
B. Incorrect: User preferences are unrelated to recommendations is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
C. Incorrect: Recommender systems cannot use data is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
D. Incorrect: Recommender systems only assign IP addresses is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
A learner is reviewing Stanford-ML-evaluation. What should they remember?
A. Correct: Model evaluation compares performance on data not used to directly fit the model is the correct answer because model evaluation compares performance on data not used to directly fit the model. Practical ML requires measuring how well a model generalizes.
B. Incorrect: Metrics are never useful in ML is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
C. Incorrect: Evaluation should only use memorized training examples is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
D. Incorrect: Generalization is unrelated to ML quality is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
When practicing Stanford Machine Learning, which option belongs under Neural Networks?
A. Incorrect: Neural networks cannot be trained from data is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
B. Incorrect: Layers and parameters never affect outputs is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
C. Incorrect: Neural networks are only physical cabling diagrams is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
D. Correct: Neural networks learn parameters across layers to model complex relationships is the correct answer because neural networks learn parameters across layers to model complex relationships. Neural networks are part of the specialization path.
When practicing Stanford Machine Learning, which option belongs under Supervised Learning?
A. Correct: Supervised learning trains models from examples that include input features and target labels is the correct answer because supervised learning trains models from examples that include input features and target labels. Supervised learning is a foundational topic in the specialization.
B. Incorrect: Supervised learning never uses labels is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
C. Incorrect: Input features are unrelated to supervised models is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
D. Incorrect: Supervised learning is only file encryption is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
A learner is reviewing Stanford-ML-unsupervised. What should they remember?
A. Incorrect: Unsupervised learning is only a keyboard shortcut is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
B. Correct: Unsupervised learning finds patterns in data without target labels is the correct answer because unsupervised learning finds patterns in data without target labels. Unsupervised learning is a core ML category.
C. Incorrect: Unsupervised learning always requires target labels is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
D. Incorrect: Pattern discovery is unrelated to unsupervised learning is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
What is the safest study takeaway for Practical Advice?
A. Incorrect: Practical ML advice means never diagnosing errors is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
B. Incorrect: Bias and variance cannot affect models is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
C. Correct: Practical ML work diagnoses bias, variance, data issues, and iteration priorities is the correct answer because practical ML work diagnoses bias, variance, data issues, and iteration priorities. The specialization includes practical ML decision-making concepts.
D. Incorrect: Iteration priorities are unrelated to model improvement is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
A learner is reviewing Stanford-ML-recommenders. What should they remember?
A. Incorrect: Recommender systems cannot use data is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
B. Correct: Recommender systems predict user preferences or item relevance from data is the correct answer because recommender systems predict user preferences or item relevance from data. Recommender systems are a common ML application area.
C. Incorrect: Recommender systems only assign IP addresses is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
D. Incorrect: User preferences are unrelated to recommendations is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
What is the safest study takeaway for Model Evaluation?
A. Correct: Model evaluation compares performance on data not used to directly fit the model is the correct answer because model evaluation compares performance on data not used to directly fit the model. Practical ML requires measuring how well a model generalizes.
B. Incorrect: Metrics are never useful in ML is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
C. Incorrect: Generalization is unrelated to ML quality is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
D. Incorrect: Evaluation should only use memorized training examples is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
What is the safest study takeaway for Neural Networks?
A. Correct: Neural networks learn parameters across layers to model complex relationships is the correct answer because neural networks learn parameters across layers to model complex relationships. Neural networks are part of the specialization path.
B. Incorrect: Neural networks cannot be trained from data is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
C. Incorrect: Neural networks are only physical cabling diagrams is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
D. Incorrect: Layers and parameters never affect outputs is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
Which answer is the best source-backed summary of this Stanford / DeepLearning.AI Machine Learning Specialization topic?
A. Incorrect: Supervised learning is only file encryption is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
B. Correct: Supervised learning trains models from examples that include input features and target labels is the correct answer because supervised learning trains models from examples that include input features and target labels. Supervised learning is a foundational topic in the specialization.
C. Incorrect: Input features are unrelated to supervised models is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
D. Incorrect: Supervised learning never uses labels is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
What is the safest study takeaway for Unsupervised Learning?
A. Incorrect: Pattern discovery is unrelated to unsupervised learning is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
B. Correct: Unsupervised learning finds patterns in data without target labels is the correct answer because unsupervised learning finds patterns in data without target labels. Unsupervised learning is a core ML category.
C. Incorrect: Unsupervised learning is only a keyboard shortcut is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
D. Incorrect: Unsupervised learning always requires target labels is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
A learner is reviewing Stanford-ML-practical. What should they remember?
A. Incorrect: Bias and variance cannot affect models is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
B. Incorrect: Iteration priorities are unrelated to model improvement is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
C. Correct: Practical ML work diagnoses bias, variance, data issues, and iteration priorities is the correct answer because practical ML work diagnoses bias, variance, data issues, and iteration priorities. The specialization includes practical ML decision-making concepts.
D. Incorrect: Practical ML advice means never diagnosing errors is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
When practicing Stanford Machine Learning, which option belongs under Recommender Systems?
A. Incorrect: Recommender systems cannot use data is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
B. Incorrect: User preferences are unrelated to recommendations is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
C. Incorrect: Recommender systems only assign IP addresses is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
D. Correct: Recommender systems predict user preferences or item relevance from data is the correct answer because recommender systems predict user preferences or item relevance from data. Recommender systems are a common ML application area.
Which statement best matches Model Evaluation for Stanford Machine Learning practice?
A. Incorrect: Generalization is unrelated to ML quality is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
B. Incorrect: Evaluation should only use memorized training examples is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
C. Incorrect: Metrics are never useful in ML is incorrect because it does not answer this stem as directly as Model evaluation compares performance on data not used to directly fit the model..
D. Correct: Model evaluation compares performance on data not used to directly fit the model is the correct answer because model evaluation compares performance on data not used to directly fit the model. Practical ML requires measuring how well a model generalizes.
Which answer is the best source-backed summary of this Stanford / DeepLearning.AI Machine Learning Specialization topic?
A. Incorrect: Neural networks cannot be trained from data is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
B. Incorrect: Neural networks are only physical cabling diagrams is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
C. Incorrect: Layers and parameters never affect outputs is incorrect because it does not answer this stem as directly as Neural networks learn parameters across layers to model complex relationships..
D. Correct: Neural networks learn parameters across layers to model complex relationships is the correct answer because neural networks learn parameters across layers to model complex relationships. Neural networks are part of the specialization path.
A learner is reviewing Stanford-ML-supervised. What should they remember?
A. Incorrect: Input features are unrelated to supervised models is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
B. Correct: Supervised learning trains models from examples that include input features and target labels is the correct answer because supervised learning trains models from examples that include input features and target labels. Supervised learning is a foundational topic in the specialization.
C. Incorrect: Supervised learning is only file encryption is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
D. Incorrect: Supervised learning never uses labels is incorrect because it does not answer this stem as directly as Supervised learning trains models from examples that include input features and target labels..
When practicing Stanford Machine Learning, which option belongs under Unsupervised Learning?
A. Incorrect: Unsupervised learning always requires target labels is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
B. Incorrect: Unsupervised learning is only a keyboard shortcut is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
C. Incorrect: Pattern discovery is unrelated to unsupervised learning is incorrect because it does not answer this stem as directly as Unsupervised learning finds patterns in data without target labels..
D. Correct: Unsupervised learning finds patterns in data without target labels is the correct answer because unsupervised learning finds patterns in data without target labels. Unsupervised learning is a core ML category.
Which answer is the best source-backed summary of this Stanford / DeepLearning.AI Machine Learning Specialization topic?
A. Incorrect: Iteration priorities are unrelated to model improvement is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
B. Incorrect: Practical ML advice means never diagnosing errors is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
C. Incorrect: Bias and variance cannot affect models is incorrect because it does not answer this stem as directly as Practical ML work diagnoses bias, variance, data issues, and iteration priorities..
D. Correct: Practical ML work diagnoses bias, variance, data issues, and iteration priorities is the correct answer because practical ML work diagnoses bias, variance, data issues, and iteration priorities. The specialization includes practical ML decision-making concepts.
Which answer is the best source-backed summary of this Stanford / DeepLearning.AI Machine Learning Specialization topic?
A. Incorrect: Recommender systems only assign IP addresses is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
B. Incorrect: Recommender systems cannot use data is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
C. Incorrect: User preferences are unrelated to recommendations is incorrect because it does not answer this stem as directly as Recommender systems predict user preferences or item relevance from data..
D. Correct: Recommender systems predict user preferences or item relevance from data is the correct answer because recommender systems predict user preferences or item relevance from data. Recommender systems are a common ML application area.
How are Stanford Machine Learning questions generated?
dotCreds builds Stanford Machine Learning practice questions from public exam objectives and DeepLearning.AI exam and documentation references. The questions are written for realistic study practice, not copied from exam dumps.
How are explanations sourced?
Each question includes an explanation and, when available, a source link back to the provider documentation or reference used to validate the answer. That keeps the practice tied to study material you can actually review.
What score do I get?
The page tracks today's answered count and accuracy for the 30-question daily set, then saves a 7-day score history on this device so you can see your recent practice trend.
Why use this site?
The site is the fastest way to start Stanford Machine Learning practice without installing anything. It is built for daily recall, quick weak-topic discovery, and source-backed explanations you can review immediately.
Why use the app when available?
The web page is the quick free sampler. If a dotCreds app is available for Stanford Machine Learning, the app is better for larger banks, focused weak-domain drills, longer review sessions, and mobile study routines.