Take the Deep Learning — Lecture 1: Introduction to ML and DL, Set 04: 20 multiple-choice questions (30 minutes) with instant scoring and a full answer…
The first 5 questions of this test, with the correct answer and explanation.
Q1. What does the lecture describe as the most important part of the promise of AI?
- A. Helping humanity move forward by assisting scientists in breakthrough discoveries across every field ✓
- B. Delivering groceries and driving cars
- C. Replacing teachers in every classroom
- D. Reducing the cost of computer hardware
Answer: A. Helping humanity move forward by assisting scientists in breakthrough discoveries across every field
Explanation: The slide says that most important, AI will help humanity move forward by assisting scientists in breakthrough discoveries across every field, from genomics to mathematics.
Q2. Under the agriculture card, which two problems are named?
- A. Soil chemistry analysis and fertiliser pricing
- B. Crop disease from leaf photos, and yield and water-stress from satellite images ✓
- C. Tractor route planning and seed distribution
- D. Livestock counting and milk-quality grading
Answer: B. Crop disease from leaf photos, and yield and water-stress from satellite images
Explanation: The agriculture card lists crop disease from leaf photos, and yield and water-stress from satellite images.
Q3. The floods and climate card names which two applications?
- A. Earthquake prediction and volcano monitoring
- B. Air-quality mapping and carbon accounting
- C. River-flood forecasting from rainfall and gauge data, and heat-wave early warning ✓
- D. Dam structural inspection and canal maintenance
Answer: C. River-flood forecasting from rainfall and gauge data, and heat-wave early warning
Explanation: The card lists river-flood forecasting from rainfall and gauge data, and heat-wave early warning.
Q4. Which languages are named as low-resource languages where speech recognition and translation are open problems?
- A. English, French and German
- B. Arabic, Persian and Turkish
- C. Hindi, Bengali and Tamil
- D. Urdu, Saraiki and Punjabi ✓
Answer: D. Urdu, Saraiki and Punjabi
Explanation: The language card names speech recognition and translation for Urdu, Saraiki, Punjabi and other low-resource languages.
Q5. The finance card lists fraud detection in mobile-wallet payments and which second problem?
- A. Credit scoring for people with no bank history ✓
- B. Stock-market index prediction
- C. Automated tax filing
- D. Currency exchange-rate arbitrage
Answer: A. Credit scoring for people with no bank history
Explanation: The finance card lists fraud detection in mobile-wallet payments and credit scoring for people with no bank history.
Take the timed test (free, no sign-up) to see your score and the answer and explanation for every question.
Q6. Which two education applications are named?
- A. Timetable optimisation and fee collection
- B. Automatic marking of handwritten scripts, and tutoring systems that adapt to each student
- C. Plagiarism detection and library cataloguing
- D. Admission screening and hostel allocation
Q7. The energy and industry card names electricity-load forecasting and what else?
- A. Designing new turbine blades
- B. Pricing electricity for domestic users
- C. Predicting machine failure before it happens
- D. Mapping transmission-line routes
Q8. Which of these is listed as an applied venue for this kind of work?
- A. NeurIPS
- B. ICLR
- C. Interspeech
- D. Elsevier Computers and Electronics in Agriculture
Q9. Which venues are named as research venues rather than applied venues?
- A. NeurIPS, ICML, ICLR and CVPR workshops
- B. IEEE Access and MDPI Sensors
- C. HEC Recognized Journals and SSRN
- D. Kaggle and Zindi
Q10. Which venues are specifically named for language and speech work?
- A. CVPR and ICCV
- B. ACL and Interspeech
- C. MDPI Sensors and Applied Sciences
- D. NeurIPS and ICML
Q11. In step 2 of the research path, what counts as a baseline?
- A. The best published result in the field
- B. A random guess at the class label
- C. Logistic regression or a small pretrained network, with the number written down
- D. The accuracy of a human expert on the same task
Q12. What does step 3 of the research path ask you to do?
- A. Change several things at once to maximise the gain
- B. Improve one thing, and measure the gain honestly
- C. Repeat the baseline on a second dataset
- D. Write the paper before running the experiment
Q13. What does step 4 say to do with your code?
- A. Keep it private until the paper is accepted
- B. Submit it with the paper as a supplementary file only
- C. Rewrite it in a different language before release
- D. Share it on GitHub
Q14. According to the rate figures cited, what is the global freelance platform median hourly rate?
- A. $19 per hour
- B. $50 per hour
- C. $125 per hour
- D. $185 per hour
Q15. What senior freelance median rate is cited, and what does the slide title say this is relative to the platform median?
- A. $50 per hour, about twice the median
- B. $185 per hour, about ten times the platform median
- C. $125 per hour, about five times the median
- D. $200 per hour, about fifteen times the median
Q16. What US average base pay is quoted for machine-learning roles?
- A. About $75,000
- B. About $95,000
- C. About $125,000
- D. About $185,000
Q17. What does the lecture say about Kaggle and Zindi competitions?
- A. They only matter for students at top universities
- B. They have replaced academic publishing entirely
- C. They require a paid subscription to enter
- D. They pay cash prizes, and a top-10% finish is a CV line recruiters recognise
Q18. What does the lecture suggest doing with Teachable Machine?
- A. Train an image classifier with your webcam: two hand gestures, 30 photos each, done in two minutes
- B. Fine-tune a large language model on your own notes
- C. Run a pretrained speech model on an Urdu voice note
- D. Write and execute a GPU notebook for the first assignment
Q19. What is suggested as a use of Hugging Face Spaces?
- A. Add and remove layers and watch a network separate spirals
- B. Run Whisper on your own Urdu voice note, or try a vision model on a photo you take
- C. Train an image classifier using your webcam
- D. Store your dataset for the semester project
Q20. What is Google Colab described as, and what will it be used for in this course?
- A. A paid cloud service, used for the final project only
- B. A local Python installation, used for all lab work
- C. A dataset repository, used to download MNIST
- D. A free GPU notebook, where the first assignment will run with no installation needed