How training data shapes AI behavior in AI for Everyone - Try It Yourself

How Training Data Shapes AI Behavior

📖 Scenario: You are learning how AI systems learn from examples. Imagine you want to teach a simple AI to recognize fruits based on their color and size.

🎯 Goal: Build a simple data set of fruits, set a rule to decide if a fruit is 'likely an apple', and apply that rule to classify the fruits.

📋 What You'll Learn

Create a dictionary called fruits with fruit names as keys and their attributes as values (color and size).

Create a variable called apple_color and set it to the string 'red'.

Use a dictionary comprehension called apple_likelihood to mark fruits as true if their color matches apple_color and size is 'medium'.

Add a final statement to count how many fruits are likely apples and store it in apple_count.

💡 Why This Matters

🌍 Real World

Understanding how AI uses training data helps in designing fair and accurate AI systems in areas like image recognition and recommendation engines.

💼 Career

This knowledge is important for roles in AI development, data science, and machine learning engineering where data quality directly affects AI behavior.

1

Create the fruit data set

Create a dictionary called fruits with these exact entries: 'banana': {'color': 'yellow', 'size': 'medium'}, 'apple1': {'color': 'red', 'size': 'medium'}, 'apple2': {'color': 'green', 'size': 'small'}, 'cherry': {'color': 'red', 'size': 'small'}.

AI for Everyone

# Create the fruits dictionary with exact entries
# Your code here

Need a hint?

Use a dictionary with fruit names as keys and another dictionary for their color and size.

2

Set the apple color to look for

Create a variable called apple_color and set it to the string 'red'.

AI for Everyone

fruits = {
    'banana': {'color': 'yellow', 'size': 'medium'},
    'apple1': {'color': 'red', 'size': 'medium'},
    'apple2': {'color': 'green', 'size': 'small'},
    'cherry': {'color': 'red', 'size': 'small'}
}
# Set apple_color to 'red'
# Your code here

Need a hint?

Assign the string 'red' to the variable apple_color.

3

Identify likely apples using a dictionary comprehension

Create a dictionary comprehension called apple_likelihood that goes through fruits.items() and sets the value to true if the fruit's color equals apple_color and its size is 'medium', otherwise false.

AI for Everyone

fruits = {
    'banana': {'color': 'yellow', 'size': 'medium'},
    'apple1': {'color': 'red', 'size': 'medium'},
    'apple2': {'color': 'green', 'size': 'small'},
    'cherry': {'color': 'red', 'size': 'small'}
}
apple_color = 'red'
# Create apple_likelihood dictionary comprehension
# Your code here

Need a hint?

Use a dictionary comprehension with for fruit, info in fruits.items() and check both color and size.

4

Count how many fruits are likely apples

Create a variable called apple_count that sums the true values in apple_likelihood.values().

AI for Everyone

fruits = {
    'banana': {'color': 'yellow', 'size': 'medium'},
    'apple1': {'color': 'red', 'size': 'medium'},
    'apple2': {'color': 'green', 'size': 'small'},
    'cherry': {'color': 'red', 'size': 'small'}
}
apple_color = 'red'
apple_likelihood = {fruit: (info['color'] == apple_color and info['size'] == 'medium') for fruit, info in fruits.items()}
# Count how many fruits are likely apples
# Your code here

Need a hint?

Use the sum() function on the values of apple_likelihood to count true entries.