SAMPLE EXAM - INTRODUCTION TO DATA SCIENCE

Code 101

PART I. Multiple-Choice Questions (8 pts)

Question 1. Underfitting is most likely caused by:
A. Models that are too simple to capture data patterns.
B. Too many training epochs.
C. Excessively complex models.
D. Oversampling minority classes.
Question 2. Which of the following is considered a machine learning performance metric?
A. Hyperparameter. B. Accuracy. C. Loss function. D. Training method.
Question 3. What is the main purpose of data labeling in machine learning?
A. To remove noise from datasets.
B. To normalize feature scales.
C. To create target variables for supervised learning.
D. To reduce dimensionality.
Question 4. Which category of machine learning uses unlabeled data?
A. Classification learning. B. Reinforcement learning.
C. Supervised learning. D. Unsupervised learning.
Question 5. Which statistical method can be used to detect data drift?
A. Cross-validation. B. Accuracy score. C. Confusion matrix. D. KS-test.
Question 6. Which of the following correctly represents the relationship among AI, Machine Learn-
ing, and Deep Learning?
A. Machine Learning ⊂ AI ⊂ Deep Learning. B. Deep Learning ⊂ Machine Learning ⊂ AI.
C. AI ⊂ Machine Learning ⊂ Deep Learning. D. All are independent fields.
Question 7. Why is ImageNet considered a milestone dataset in deep learning?
A. Because it requires no computation resources.
B. Because its large-scale labeled data enabled the breakthrough of deep CNNs.
C. Because it has only a few categories.
D. Because it is only used for handwritten recognition.
Question 8. Which dataset is commonly used to predict real estate prices?
A. COCO. B. CIFAR-10. C. ImageNet. D. Boston House Price.
Question 9. What is a key difference between MNIST and CIFAR-10?
A. MNIST is grayscale while CIFAR-10 is colored.
B. Both have colored images.
C. CIFAR-10 has fewer images than MNIST.
D. MNIST has more categories than CIFAR-10.
Question 10. Cloud computing primarily supports Digital Transformation by enabling:
A. Offline data analysis.
B. Manual data entry.
C. Centralized data storage and scalable computing.
D. Hardware dependency.
Question 11. Which of the following is NOT a key aspect of Digital Transformation?
A. Process re-engineering. B. Cultural change.
C. Technology adoption. D. Hardware replacement only.

Final Exam - EP15.DSAI1030 - Fall 2025 Page 1/5 Code 101

Question 12. Netflix Prize is associated with which application?
A. Route optimization. B. Recommender systems.
C. Fraud detection. D. Medical imaging.
Question 13. What is the relationship between Data Mining and Data Science?
A. Data Mining is a subset of Data Science. B. Data Science is a subset of Data Mining.
C. They are completely unrelated. D. They are identical.
Question 14. Dashboards and AI tools that help doctors plan treatment represent:
A. Personalization. B. Fraud detection.
C. Decision Support Systems. D. Optimization.
Question 15. Which company used data science to stock Pop-Tarts before hurricanes?
A. Target. B. Amazon. C. Walmart. D. Costco.
Question 16. Amazon recommendations mainly rely on:
A. Browsing and purchase history. B. TV advertisements.
C. Employee suggestions. D. Random guessing.
Question 17. What is the main goal of Data Science?
A. To turn raw data into actionable knowledge. B. To design databases.
C. To build mobile apps. D. To manage cloud services.
Question 18. Which three key components form Data Science?
A. Mathematics, Communication, Finance. B. Business, Marketing, Networking.
C. Statistics, Computing, Domain Expertise. D. Programming, Accounting, Visualization.
Question 19. Which AWS service allows SQL queries directly on data stored in S3 without loading
it into a database?
A. Amazon DynamoDB. B. Amazon Athena.
C. Amazon EMR. D. Amazon Redshift.
Question 20. Which combination correctly matches a Big Data processing type with its technol-
ogy?
A. Batch processing – Apache Flink. B. Visualization – Apache Kafka.
C. Streaming processing – Apache Storm. D. Storage – Apache Spark.
Question 21. In an AWS Big Data workflow, which service is mainly responsible for ETL (Extract,
Transform, Load)?
A. Amazon S3. B. Amazon DynamoDB.
C. AWS Glue. D. Amazon QuickSight.

Question 22. Which tool is mainly used for batch processing of large-scale data?
A. Redis. B. Elasticsearch. C. Apache Spark. D. MongoDB.

Question 23. Which technology is commonly used for streaming data ingestion?
A. Tableau. B. Hive. C. Apache Kafka. D. Hadoop HDFS.

Question 24. Which combination correctly matches a Big Data processing type with its technol-
ogy?
A. Batch processing – Apache Flink. B. Storage – Apache Spark.
C. Visualization – Apache Kafka. D. Streaming processing – Apache Storm.

Question 25. Which type of database typically prioritizes ACID properties?

A. NoSQL databases. B. Graph databases.
C. Key-value databases. D. SQL (relational) databases.

Final Exam - EP15.DSAI1030 - Fall 2025 Page 2/5 Code 101

Question 26. Which of the following is considered a soft skill for data science professionals?
A. Hypothesis testing. B. ETL processing.
C. Gradient descent. D. Storytelling with data.

Question 27. What is one primary goal of a data science career?

A. Transforming raw data into actionable insights.
B. Designing computer hardware.
C. Managing company finances.
D. Developing mobile applications.

Question 28. Which of the following skill sets is most comprehensive for a Data Scientist?
A. Cloud deployment and system administration.
B. SQL, Excel, reporting.
C. Python, Tableau, communication.
D. Statistics, Python/R, machine learning, data engineering, domain knowledge.

Question 29. Which of the following is an example of output (label) in the Melbourne housing case
study?
A. Property type. B. Number of rooms.
C. House price. D. Distance to city center.

Question 30. Which of the following models is listed as an advanced regression technique?
A. Naive Bayes. B. k-Means. C. Gradient Boosting. D. PCA.

Question 31. Which step comes immediately after Data Collection in the Data Science Process?
A. Preprocessing. B. Evaluation.
C. Understanding the Problem and Data. D. Deployment.

Question 32. Which of the following is a deep learning approach mentioned for time-series forecast-
ing?
A. Logistic Regression. B. Decision Tree. C. Naive Forecasting. D. LSTM.

Question 33. One key disadvantage of K-Nearest Neighbors (KNN) is:

A. It only works with text data.
B. It requires heavy memory and is slow at prediction.
C. It cannot handle classification.
D. It has a complex training phase.

Question 34. In deployment and monitoring, what should be done when model performance de-
grades due to data drift?
A. Delete the model. B. Ignore the issue.
C. Retrain the model. D. Stop predictions permanently.

Question 35. Which of the following is a classification problem?

A. Predicting stock prices.
B. Predicting exam grades.
C. Predicting whether a customer will churn or not.
D. Predicting house rental price.

Final Exam - EP15.DSAI1030 - Fall 2025 Page 3/5 Code 101

Question 36. Which of the following is NOT a key component of a Data Ecosystem?
A. Data analytics. B. Data compression. C. Data storage. D. Data ingestion.

Question 37. Which characteristic best describes a Data Warehouse?

A. Stores unstructured data such as images and videos.
B. Schema-on-read and flexible data storage.
C. Optimized for machine learning on raw data.
D. Schema-on-write with structured, curated data.

Question 38. In a modern data ecosystem, what is the primary role of a Data Hub?
A. Integrate, standardize, and distribute data across systems.
B. Store large volumes of raw data.
C. Perform machine learning model training.
D. Replace the data warehouse.

Question 39. Which of the following is NOT a visual element commonly used in data visualiza-
tion?
A. Tables. B. Operating systems. C. Graphs. D. Charts.

Question 40. Which plot shows the five-number summary (min, Q1, median, Q3, max)?
A. Line chart. B. Density plot. C. Histogram. D. Box plot.

Question 41. Which visualization is best for hierarchically structured data?

A. Line chart. B. Treemap. C. Donut chart. D. Histogram.

Question 42. You want to compare distributions of exam scores across three different classes. The
best visualization is:
A. Violin plot or box plot. B. Scatter plot.
C. Single pie chart. D. Histogram with one bin.

Question 43. A KDE (Kernel Density Estimation) plot improves upon a histogram primarily by:
A. Providing a smoother estimate of the data distribution.
B. Making the visualization 3D.
C. Removing the need for data.
D. Automatically increasing number of bins.

Question 44. A bar chart is most appropriate when you want to:
A. Display time-series only. B. Visualize geographic maps.
C. Show probability density. D. Display categorical data values.

Question 45. Which of the following is NOT a visual element commonly used in data visualiza-
tion?
A. Operating systems. B. Charts. C. Graphs. D. Tables.

Question 46. Which plot shows the five-number summary (min, Q1, median, Q3, max)?
A. Box plot. B. Histogram. C. Line chart. D. Density plot.

Final Exam - EP15.DSAI1030 - Fall 2025 Page 4/5 Code 101

Question 47. Which visualization is best for hierarchically structured data?
A. Donut chart. B. Histogram. C. Line chart. D. Treemap.

Question 48. Data visualization is primarily used to:

A. To decorate reports with graphics.
B. To represent data using visual elements for better understanding.
C. To remove the need for raw data interpretation.
D. To replace statistical analysis.

Question 49. Which step is included in the data visualization process?

A. Data encryption.
B. Building neural networks.
C. Removing missing values only.
D. Determining what insights can be drawn from the data.

Question 50. You want to compare distributions of exam scores across three different classes. The
best visualization is:
A. Single pie chart. B. Violin plot or box plot.
C. Histogram with one bin. D. Scatter plot.

PART II. Constructed-Response Questions (2 pts)

Question 51. What is the overfitting in machine learning? What can you observe when the overfitting
occurs? What are the reasons behind overfitting? What should you do to address this issue?
- - - - - - - - - - END- - - - - - - - - -

Final Exam - EP15.DSAI1030 - Fall 2025 Page 5/5 Code 101