Bob's AI Fundamentals Platform

Question 1

A retail company wants to predict which customers are likely to stop purchasing in the next three months. Which machine learning approach is most appropriate?

A) Reinforcement learning B) Supervised learning C) Unsupervised learning D) Generative AI

Question 2

A team has thousands of customer records but no labels indicating customer types. They want to identify natural groupings within the data. Which technique would be most appropriate?

A) Classification B) Regression C) Clustering D) Forecasting

Question 3

A machine learning model achieves 99% accuracy during training but performs poorly on new data. What is the most likely explanation?

A) The model is underfitting B) The dataset is too small to train C) The model is overfitting D) The model has too few features

Question 4

Which statement best describes the purpose of a training dataset?

A) To evaluate the final model after deployment B) To provide examples the model learns from C) To store predictions generated by the model D) To monitor model performance over time

Question 5

A bank uses machine learning to predict whether a loan applicant is likely to repay a loan. What is the predicted outcome commonly called?

A) A feature B) A label C) A parameter D) A cluster

Question 6

A company wants to estimate next month's sales revenue based on historical sales data. Which machine learning task is most appropriate?

A) Clustering B) Classification C) Recommendation D) Regression

Question 7

Which statement best explains why data quality is important in machine learning?

A) Poor-quality data can lead to unreliable predictions B) Data quality only affects training speed C) Data quality matters only for large organizations D) Models automatically correct poor-quality data

Question 8

A model predicts whether emails are spam or legitimate. This is an example of:

A) Regression B) Classification C) Clustering D) Forecasting

Question 9

A project team wants to measure how well a machine learning model performs on data it has never seen before. What should they use?

A) Training data B) Synthetic data only C) Test data D) Randomly generated predictions

Question 10

An organization is selecting a machine learning model for a customer-facing application. Beyond accuracy, which consideration is often important for business adoption?

A) The model's file size only B) The number of developers involved C) Explainability and trustworthiness D) The programming language used

Question 11

A healthcare organization develops a machine learning model to identify patients at risk of missing appointments. Why is model bias a concern?

A) Biased models may unfairly impact certain groups of patients B) Bias always improves prediction accuracy C) Bias only affects image-based models D) Bias eliminates the need for human oversight

Question 12

A model performs poorly on both training data and new data. What is the most likely issue?

A) Overfitting B) Data leakage C) Underfitting D) Model drift

Question 13

A retailer wants to recommend products to customers based on previous purchasing behavior. Which machine learning application is being used?

A) Fraud detection B) Recommendation systems C) Sentiment analysis D) Forecasting

Question 14

A machine learning team separates data into training and test datasets. What is the primary purpose of the test dataset?

A) To increase the amount of training data B) To store production predictions C) To evaluate model performance objectively D) To create additional labels

Question 15

A company notices that customer behavior has changed significantly since a model was originally trained. As a result, prediction accuracy declines. What is this situation commonly called?

A) Clustering B) Data normalization C) Feature engineering D) Model drift

Question 16

A logistics company wants to estimate delivery times based on distance, traffic conditions, and weather. Which type of input information are distance, traffic, and weather considered?

A) Labels B) Features C) Predictions D) Metrics

Question 17

A model correctly identifies 95% of legitimate transactions but misses many fraudulent transactions. Which metric would help reveal this weakness?

A) Recall B) Storage utilization C) Dataset size D) Processing speed

Question 18

A company wants to understand which factors most influence a model's predictions before deploying it in a regulated industry. What is the primary concern?

A) Increasing GPU usage B) Improving data storage capacity C) Model explainability D) Reducing internet bandwidth

Question 19

A machine learning project begins by collecting customer records from several systems. What should the team do before training a model?

A) Immediately deploy the model B) Remove all historical records C) Clean and prepare the data D) Ignore incomplete records entirely

Question 20

A company evaluates two models. Model A achieves slightly higher accuracy, while Model B is easier to explain and audit. For a highly regulated industry, which choice may be preferable?

A) Model A, because accuracy is the only consideration B) Model B, because transparency may be required for compliance C) Neither model should be deployed D) Both models should always be combined

Question 21

A financial institution uses machine learning to approve loan applications. Regulators require explanations for decisions made by the model. Which approach best supports this requirement?

A) Use the most complex model available regardless of transparency B) Deploy the model without documenting its behavior C) Prioritize explainable models and maintain decision records D) Hide model outputs from reviewers

Question 22

A company develops a fraud detection model where fraudulent transactions represent only 1% of all transactions. Which challenge is most likely to arise?

A) The model may appear accurate while missing many fraud cases B) The model will automatically become unbiased C) The model will require less training data D) The model will only produce false positives

Question 23

A machine learning team accidentally includes future information in the training dataset that would not be available when making real-world predictions. What issue has occurred?

A) Feature engineering B) Data leakage C) Clustering D) Normalization

Question 24

An organization wants to ensure a machine learning model remains effective after deployment. Which practice is most important?

A) Training the model once and never reviewing it again B) Monitoring performance and retraining when necessary C) Preventing users from providing feedback D) Reducing the amount of collected data over time

Question 25

A business leader asks why machine learning predictions should not always be treated as facts. What is the best response?

A) Machine learning systems generate estimates based on patterns in data and can still make mistakes B) Machine learning systems always provide guaranteed outcomes C) Machine learning systems eliminate uncertainty from decisions D) Machine learning systems do not rely on data

Machine Learning Quiz