Name: Date:
How Machines Learn: End-of-Topic Test
Instructions
📢 Time Allowed: 40 Minutes
Equipment: Pen, pencil, eraser.
Instructions:
● Answer ALL questions.
● Write your answers in the spaces provided.
● Read each question carefully before starting.
Section A: Key Concepts
Select the correct answer for each question. These questions test your understanding of the
fundamental definitions of Artificial Intelligence and Machine Learning.
1. Which statement best a) Writing strict rules for a
describes Machine computer to follow
Learning? step-by-step.
b) A system where
computers learn from data
to find patterns without
explicit programming.
c) A physical robot that can
move objects but cannot
think.
d) The process of fixing
broken computer hardware.
2. How does Machine a) Traditional programming
Learning differ from uses data to make rules; ML
traditional uses rules to make data.
programming? b) Traditional programming
is faster than Machine
Learning.
c) Traditional programming
gives computers examples;
ML gives computers rules.
d) Traditional programming
relies on specific rules (If X,
then Y); ML learns from
examples.
Vocabulary Match
Match the following terms to their correct definitions.
Terms: Training, Model, Feedback, Input
Definition Word
1. The data given to a machine to help it learn (e.g., photos of
messy rooms).
2. The 'engine' or system created by the computer after it has
learned from data.
3. Information given to the machine about whether its
prediction was right or wrong.
4. The process of teaching a machine to recognize patterns.
Section B: The Learning Process
Machines follow a specific cycle to learn and improve. Study the diagram below and then
complete the paragraph describing the process.
Word bank: prediction, accuracy, input, training, feedback, model
The machine learning cycle begins with __________ data, such as a large collection of images.
During the __________ phase, the computer analyzes this data to find patterns. It builds a
__________ which it uses to make a guess about new data. This guess is called a __________.
Finally, the system receives __________ to tell it if it was right or wrong, which helps improve
its __________ over time.
Section C: Types of Machine Learning
Identify the three main types of machine learning: Supervised, Unsupervised, and
Reinforcement.
Word bank: Supervised Learning, Unsupervised Learning,
Reinforcement Learning
1. A system is given labeled photos of camels and goats to
learn how to tell them apart. This is __________.
2. A robot learns to navigate a maze by trying different paths;
it gets a point for finding the exit and loses a point for
hitting a wall. This is __________.
3. A computer is given a list of random customer data and
groups them based on similarities without any instructions.
This is __________.
Real-World Application
For each scenario below, choose the correct learning type from the options provided.
1. Self-driving cars learn to drive safely a) Supervised Learning
by receiving a 'penalty' in their code if b) Unsupervised Learning
they get too close to another object. c) Reinforcement Learning
2. Netflix recommends movies by a) Supervised Learning
grouping you with other users who b) Unsupervised Learning
watch similar things, without c) Reinforcement Learning
knowing exactly who you are.
3. A smart fridge scanner is trained to a) Supervised Learning
recognize 'apple', 'banana', and b) Unsupervised Learning
'orange' by scanning thousands of c) Reinforcement Learning
items with price tags.
Section D: Analysis & Critical Thinking
Answer the following questions in full sentences. Focus on explaining why and how.
1. The 'Child and the Toys' Analogy
Imagine a child is given a box of mixed toys. Explain the difference between Supervised and
Unsupervised learning using this example.
2. Designing a Cleaning Robot
Zayed is designing a robot to organize his bedroom. The robot needs to learn where to put school
books versus where to put clothes.
A) Which type of machine learning would be best for this task?
B) Explain how the robot would use feedback to get better at this task over time.
Answer Key
Section A: Key Concepts
Multiple Choice:
1. A system where computers learn from data to find patterns without explicit programming.
2. Traditional programming relies on specific rules (If X, then Y); ML learns from examples.
Vocabulary Practice:
1. Input
2. Model
3. Feedback
4. Training
Section B: The Learning Process
Cloze:
The machine learning cycle begins with input data, such as a large collection of images. During
the training phase, the computer analyzes this data to find patterns. It builds a model which it
uses to make a guess about new data. This guess is called a prediction. Finally, the system
receives feedback to tell it if it was right or wrong, which helps improve its accuracy over time.
Section C: Types of Machine Learning
Gap Fills:
1. A system is given labeled photos of camels and goats to learn how to tell them apart. This is
Supervised Learning.
2. A robot learns to navigate a maze by trying different paths; it gets a point for finding the exit
and loses a point for hitting a wall. This is Reinforcement Learning.
3. A computer is given a list of random customer data and groups them based on similarities
without any instructions. This is Unsupervised Learning.
Multiple Choice:
1. Reinforcement Learning
2. Unsupervised Learning
3. Supervised Learning
Section D: Analysis & Critical Thinking
Answer:
In supervised learning, a teacher would sit with the child and tell them 'This is a car' or 'This is a
block', and correct them if they are wrong. In unsupervised learning, the child is left alone with the
toys and decides to organize them their own way, perhaps grouping them by color (red toys vs
blue toys) or size, without being told the correct groups.
Answer:
A) Supervised learning (or Reinforcement learning) would be best.
B) If the robot puts a shirt on the bookshelf, Zayed gives it negative feedback (or tells it 'wrong
location'). The robot updates its model so it is less likely to make that mistake again. If it puts the
shirt in the closet, it gets positive feedback, reinforcing the correct behavior.