These slides are designed to be projected or used as a visual guide during instruction.
You can copy the content into your preferred presentation software (PowerPoint, Google Slides, etc.).

AI and Mathematics:
Pattern Recognition & Predictive Modeling

How AI Uses the Same Math You're Learning in Class
Grades 6-8 | Mathematics & AI | Evolve AI Institute
1

The Prediction Challenge

Look at these average monthly high temperatures:

Jan: 33°F    Feb: 37°F    Mar: 48°F    Apr: 62°F    May: 73°F    Jun: 82°F
What do you predict for July?
Write your prediction on a sticky note. Don't share yet!
The AI predicted: 87°F
The actual temperature: 86°F
How close was your prediction? What strategy did your brain use?
2

Today's Big Idea

AI uses the SAME math
you're learning in class.
Pattern recognition. Mean, median, mode.
Scatter plots. Trend lines. Predictions.
The more math you know, the better you understand how AI works --
and the better equipped you are to build and evaluate AI systems.
3

What Is Pattern Recognition?

Pattern recognition is the ability to detect regularities, rules, or trends in data.

Your Brain Does It:

  • Recognizing a friend's face in a crowd
  • Knowing that "i before e except after c"
  • Predicting the next number: 2, 4, 6, 8, ...
  • Expecting rain when you see dark clouds

AI Does It Too:

  • Identifying faces in photos
  • Autocorrecting your spelling
  • Predicting stock market trends
  • Forecasting tomorrow's weather
Key Difference: You can spot patterns in small amounts of data. AI can find patterns in millions of data points -- but it uses the same mathematical principles.
4

Finding the Rule: Number Sequences

Can you figure out the rule and predict the next number?

Sequence A:   2,   6,   18,   54,   ???
Rule: Multiply by 3 each time.   Next: 162
Sequence B:   3,   7,   11,   15,   ???
Rule: Add 4 each time.   Next: 19
Sequence C:   100,   90,   81,   73,   ???
Rule: Subtract 10, then 9, then 8, then...   Next: 66
Think: When Netflix recommends a show, it looks at sequences in your viewing history. What "pattern rule" might it find?
5

Mean, Median, Mode: What's "Normal"?

AI needs to know what is typical in data. These three measures help:

Test Scores:   78,   85,   92,   85,   70,   88,   95,   85,   80,   82

Calculating:

Mean = 840 / 10 = 84

Median = (82 + 85) / 2 = 83.5

Mode = 85 (appears 3 times)

When to Use Each:

Mean: Best when no outliers

Median: Best when outliers exist

Mode: Best for categories or "most popular"

AI Example: A spam filter calculates the mean number of exclamation marks in known spam emails (12 on average). If a new email has 15 exclamation marks, the AI flags it as likely spam.
6

Outlier Alert: When the Mean Lies

What if one student scored very differently?

78,   85,   92,   85,   70,   88,   95,   85,   80,   15

New Mean

77.3

Dropped from 84 to 77.3!

New Median

82.5

Only dropped from 83.5 to 82.5

Key Insight: The median is more "resistant" to outliers. AI systems must decide which measure to use depending on the data. Smart AI programmers use the median when outliers are present -- just like smart math students do!
7

Scatter Plots: Seeing Relationships

A scatter plot shows the relationship between two variables.

Hours of Practice1235710
Free Throw Accuracy %303545557080
Steps to create a scatter plot:
  1. Label the x-axis (Hours of Practice) and y-axis (Accuracy %)
  2. Choose appropriate scales for each axis
  3. Plot each data point as a dot
  4. Look for the overall pattern: Is it going up? Down? No pattern?
Prediction Question: Using the pattern, if a player practices 8 hours, what accuracy would you predict? (Think: look at the pattern between 7 and 10 hours.)
8

Trend Lines: The Key to Prediction

A trend line (line of best fit) summarizes the overall direction of the data.

How to Draw a Good Trend Line:

  • Use a ruler -- it must be straight!
  • Follow the general direction of the data
  • Try to have roughly equal points above and below the line
  • It does NOT need to pass through any specific data point

How to Use It for Prediction:

  • Find your x-value on the horizontal axis
  • Draw a line up to the trend line
  • Draw a line across to the y-axis
  • Read the predicted y-value
AI Connection: This is exactly what AI "linear regression" does -- except AI uses a precise mathematical formula to find the best possible line, while we draw it by eye. Same math, different precision!
9

Real-World AI Predictions Powered by Math

Weather Forecasting

AI analyzes patterns in decades of temperature, humidity, pressure, and wind data. It builds trend models for each variable and combines them to predict tomorrow's weather. It uses the same scatter plots and trend analysis you're learning today -- just with billions more data points.

Recommendation Systems

Netflix, Spotify, and YouTube analyze patterns in what you watch, listen to, and click. They calculate the "mean preferences" of users like you and predict what you'll enjoy next. If users who like songs A and B also tend to like song C, the AI predicts you'll like C too.

Sports Analytics

NBA teams use AI to analyze player statistics -- points, rebounds, assists -- and predict future performance. They create scatter plots of age vs. performance to predict when a player will peak. Sound familiar?

Medical Diagnosis

AI detects patterns in medical data (blood test results, vital signs, imaging) by comparing a patient's numbers to the mean values for healthy vs. sick populations. Pattern recognition helps catch diseases earlier.

10

How Good Is Our Prediction?

Scientists and AI developers use percent error to measure prediction accuracy.

Percent Error = |Predicted - Actual| ÷ Actual × 100

Example:

You predicted 68 ice cream bars sold. The actual number was 65.

Percent Error = |68 - 65| ÷ 65 × 100 = 3 ÷ 65 × 100 = 4.6%

That means your prediction was only 4.6% off -- great job!

0-10% Error

Excellent prediction!

10-25% Error

Good estimate, room to improve

11

Activity Time: Ice Cream Scatter Plot

Let's build a scatter plot together using real cafeteria data!

Temperature (°F)657075808590
Ice Cream Bars Sold182534405258
Step-by-step:
  1. Label x-axis: "Temperature (°F)" -- scale from 60 to 100
  2. Label y-axis: "Ice Cream Bars Sold" -- scale from 0 to 70
  3. Plot each of the 6 data points
  4. Draw your trend line through the middle of the points
  5. Predict: How many bars sell at 95°F?
Approximate slope: (58 - 18) / (90 - 65) = 40 / 25 = 1.6 bars per degree
That means for every 1°F increase, about 1.6 more ice cream bars sell!
12

Data Detective Challenge

🔍

Time to become Data Detectives!

Your Mission:

  • Analyze real-world data
  • Find patterns and trends
  • Calculate statistics (mean, median, mode)
  • Make predictions
  • Present your findings

Group Roles:

  • Data Manager: Organizes the dataset
  • Calculator: Does the math
  • Grapher: Creates visuals
  • Presenter: Explains findings
You have 25 minutes. Each group will present for 2-3 minutes.
After presentations, we'll compare your predictions to the AI's predictions!
13

Human vs. AI: How Did We Compare?

Scenario Actual Value Class Prediction AI Prediction Who Was Closer?
Basketball Points 22 (fill in) 21.4
Pizza Orders 175 (fill in) 172
Rainfall 3.4 in. (fill in) 3.2 in.

AI's Advantage:

  • Processes millions of data points
  • Never gets tired or distracted
  • Calculates in milliseconds
  • Can consider many variables at once

Human's Advantage:

  • Common sense and context
  • Creative problem-solving
  • Can ask "Does this make sense?"
  • Understands factors not in the data
14

What We Learned Today

  • ✓ AI uses pattern recognition -- the same skill you use with number sequences
  • ✓ AI calculates mean, median, and mode to understand data -- just like you did today
  • ✓ AI draws trend lines through scatter plots to make predictions -- exactly what you practiced
  • Percent error measures how accurate a prediction is -- whether made by a human or an AI
  • ✓ The math you learn in class IS the foundation of AI
15

You Can Think Like an AI!

"Every time you spot a pattern, calculate an average, or draw a trend line, you're doing the exact same math that powers the AI tools you use every day."
The more math you learn, the better you understand --
and can shape -- the AI-powered world around you.
Evolve AI Institute | Free AI Education for All
evolveaiinstitute.com
16