Solved Exercises on Modeling with Functions in Algebra 2

Master choosing appropriate models: linear, quadratic, exponential, logarithmic, and trigonometric functions for real-world scenarios.

Solution: Exercises 1 to 3
1 Linear growth model
Exercise 1
A plant grows 2.5 cm per week. When planted, it was 5 cm tall. Write a linear function to model the height of the plant over time and predict its height after 8 weeks.
Definition:

Linear function: f(x) = mx + b, where m is the rate of change and b is the initial value.

Modeling method:
  1. Identify the independent variable (time)
  2. Identify the dependent variable (height)
  3. Determine the rate of change (slope)
  4. Determine the initial value (y-intercept)
  5. Write the function
Rate
2.5 cm/week
Initial
5 cm
Function
h(t) = 2.5t + 5
Step 1: Identify variables

Independent variable: t (weeks)

Dependent variable: h (height in cm)

Step 2: Identify rate of change

The plant grows at a constant rate of 2.5 cm per week

So m = 2.5 (slope)

Step 3: Identify initial value

When t = 0, h = 5 cm

So b = 5 (y-intercept)

Step 4: Write the linear function

h(t) = 2.5t + 5

Step 5: Predict height after 8 weeks

h(8) = 2.5(8) + 5 = 20 + 5 = 25 cm

h(t) = 2.5t + 5
Height after 8 weeks: 25 cm
Final answer:

The linear model is h(t) = 2.5t + 5, and the plant will be 25 cm tall after 8 weeks.

Applied rules:

• Linear model: Use when rate of change is constant

• Rate of change: Slope represents the constant rate

• Initial value: Y-intercept represents starting condition

2 Quadratic motion model
Exercise 2
A ball is thrown upward from a height of 3 meters with an initial velocity of 15 m/s. Model the height of the ball using a quadratic function and find the maximum height reached.
Definition:

Projectile motion: h(t) = -½gt² + v₀t + h₀, where g = 9.8 m/s², v₀ is initial velocity, h₀ is initial height.

Given
v₀ = 15 m/s, h₀ = 3 m
Function
h(t) = -4.9t² + 15t + 3
Max Height
14.5 m
Step 1: Write the quadratic function

h(t) = -½gt² + v₀t + h₀

h(t) = -½(9.8)t² + 15t + 3

h(t) = -4.9t² + 15t + 3

Step 2: Find the vertex (maximum point)

For a quadratic in form ax² + bx + c, vertex occurs at x = -b/(2a)

t = -15/(2(-4.9)) = -15/(-9.8) = 15/9.8 ≈ 1.53 seconds

Step 3: Calculate maximum height

h(1.53) = -4.9(1.53)² + 15(1.53) + 3

h(1.53) = -4.9(2.34) + 22.95 + 3

h(1.53) = -11.47 + 22.95 + 3 = 14.48 ≈ 14.5 meters

Step 4: Verify the result

The quadratic model is appropriate because gravity causes constant acceleration, resulting in parabolic motion.

h(t) = -4.9t² + 15t + 3
Max height: 14.5 m at t = 1.53 s
Final answer:

The quadratic model is h(t) = -4.9t² + 15t + 3, and the ball reaches a maximum height of approximately 14.5 meters.

Applied rules:

• Quadratic model: Use for constant acceleration scenarios

• Vertex formula: t = -b/(2a) for maximum/minimum

• Physics equation: h(t) = -½gt² + v₀t + h₀

3 Exponential decay model
Exercise 3
A radioactive substance has a half-life of 10 years. If there are initially 200 grams, write an exponential function to model the remaining amount and find how much remains after 30 years.
Definition:

Exponential decay: A(t) = A₀(1/2)^(t/T), where A₀ is initial amount, T is half-life, and t is time.

Given
A₀ = 200g, T = 10 years
Function
A(t) = 200(1/2)^(t/10)
After 30 years
25 grams
Step 1: Identify the exponential decay model

For half-life problems, use A(t) = A₀(1/2)^(t/T)

Where: A₀ = 200g, T = 10 years

Step 2: Write the exponential function

A(t) = 200(1/2)^(t/10)

Step 3: Calculate remaining amount after 30 years

A(30) = 200(1/2)^(30/10) = 200(1/2)³ = 200(1/8) = 25 grams

Step 4: Verify the result

After 10 years: 200 → 100 (half)

After 20 years: 100 → 50 (half again)

After 30 years: 50 → 25 (half again) ✓

A(t) = 200(1/2)^(t/10)
Amount after 30 years: 25 grams
Final answer:

The exponential model is A(t) = 200(1/2)^(t/10), and 25 grams remain after 30 years.

Applied rules:

• Exponential model: Use for percentage-based growth/decay

• Half-life: Amount reduces by 50% every half-life period

• General form: A(t) = A₀ · b^(t/k) for exponential processes

Modeling Fundamentals
f(x) = mx + b
Linear Model
Linear
mx + b
Constant rate
Quadratic
ax² + bx + c
Changing rate
Exponential
abˣ
Percentage change
Logarithmic
a + b ln(x)
Decreasing growth
Key definitions:

Mathematical model: A mathematical representation of a real-world situation

Rate of change: How one quantity changes with respect to another

Regression: Statistical method to find the best-fitting model

Model Selection: Linear (constant rate), Quadratic (changing rate), Exponential (percentage change), Logarithmic (decreasing growth)
Goodness of Fit: Correlation coefficient, residual analysis, visual inspection
Tip 1: Look at the pattern of change in your data to determine the appropriate model type.
Tip 2: Linear models have constant first differences, exponential models have constant ratios.
Tip 3: Quadratic models have constant second differences.
Tip 4: Always check if your model makes sense in the real-world context.
Solution: Exercises 4 to 5
4 Logarithmic growth model
Exercise 4
The learning curve for typing speed can be modeled by S(t) = 40 - 30e^(-0.1t), where S is words per minute and t is weeks of practice. Find the maximum possible typing speed and how long it takes to reach 30 wpm.
Definition:

Logarithmic/exponential growth: Models where the rate of growth decreases over time, approaching a limiting value.

Function
S(t) = 40 - 30e^(-0.1t)
Limit
40 wpm
Time for 30 wpm
≈ 6.93 weeks
Step 1: Find the maximum possible typing speed

As t → ∞, e^(-0.1t) → 0

Therefore, S(t) → 40 - 30(0) = 40 wpm

Step 2: Find when speed reaches 30 wpm

30 = 40 - 30e^(-0.1t)

30e^(-0.1t) = 40 - 30 = 10

e^(-0.1t) = 10/30 = 1/3

Step 3: Solve for t

ln(e^(-0.1t)) = ln(1/3)

-0.1t = ln(1/3) = -ln(3)

t = ln(3)/0.1 ≈ 1.099/0.1 ≈ 10.99 weeks

Step 4: Verify the result

S(10.99) = 40 - 30e^(-0.1×10.99) = 40 - 30e^(-1.099) = 40 - 30(1/3) = 40 - 10 = 30 ✓

Max speed: 40 wpm
Time for 30 wpm: ~11 weeks
Final answer:

The maximum possible typing speed is 40 wpm, and it takes approximately 11 weeks to reach 30 wpm.

Applied rules:

• Exponential decay: e^(-kt) approaches 0 as t increases

• Natural logarithm: ln(e^x) = x

• Limiting value: Maximum achievable value as t approaches infinity

5 Model comparison
Exercise 5
Given data points: (1, 5), (2, 12), (3, 25), (4, 44), (5, 69), determine which model (linear, quadratic, or exponential) best fits the data and justify your choice.
Definition:

Model selection: Analyze first differences (linear), second differences (quadratic), and ratios (exponential) to determine the best model.

Data
(1,5), (2,12), (3,25), (4,44), (5,69)
First diffs
7, 13, 19, 25
Second diffs
6, 6, 6
Step 1: Calculate first differences

12 - 5 = 7

25 - 12 = 13

44 - 25 = 19

69 - 44 = 25

First differences: 7, 13, 19, 25 (not constant)

Step 2: Calculate second differences

13 - 7 = 6

19 - 13 = 6

25 - 19 = 6

Second differences: 6, 6, 6 (constant)

Step 3: Calculate ratios (for exponential check)

12/5 = 2.4

25/12 ≈ 2.08

44/25 = 1.76

69/44 ≈ 1.57

Ratios are not constant

Step 4: Determine the model

Since second differences are constant, the data follows a quadratic model.

Quadratic model best fits
Second differences: 6, 6, 6
Final answer:

A quadratic model best fits the data because the second differences are constant (6, 6, 6).

Applied rules:

• Linear model: Constant first differences

• Quadratic model: Constant second differences

• Exponential model: Constant ratios

• Model identification: Use differences and ratios to determine pattern

Detailed Summary: Choosing Appropriate Models
f(x) = ax^n + ... + c
Polynomial Models
Key definitions:

Linear model: f(x) = mx + b, constant rate of change

Quadratic model: f(x) = ax² + bx + c, changing rate of change

Exponential model: f(x) = abˣ, percentage rate of change

Logarithmic model: f(x) = a + b ln(x), decreasing rate of change

Model Selection Methodology:
  1. Plot the data: Visual inspection often reveals the pattern
  2. Calculate differences: First differences (linear), second differences (quadratic)
  3. Calculate ratios: For exponential models
  4. Consider the context: What type of process is occurring?
  5. Test the model: Verify with additional data points
Tip 1: Linear: Look for constant first differences, steady growth/decline.
Tip 2: Quadratic: Look for constant second differences, parabolic shape.
Tip 3: Exponential: Look for constant ratios, rapid growth/decay.
Tip 4: Logarithmic: Look for growth that slows over time, approaching a limit.
Common errors: Misidentifying patterns, ignoring context, using inappropriate domain restrictions.
Exam preparation: Practice identifying patterns, memorize characteristic behaviors of each function type.
Model identification criteria:
Model Type Pattern in Data Real-world Examples
Linear Constant first differences Constant speed, hourly wages
Quadratic Constant second differences Projectile motion, area problems
Exponential Constant ratios Population growth, compound interest
Logarithmic Decreasing differences Learning curves, pH scale
Model Comparison: Different Growth Patterns
Exercise 6: Comparing Model Types
Compare linear, quadratic, and exponential growth: f(x) = 2x + 5, g(x) = x² + 3, h(x) = 2^x + 1

Analysis: The chart shows how different models grow at different rates over time.

  • Linear: Steady, constant growth rate
  • Quadratic: Accelerating growth rate
  • Exponential: Rapidly accelerating growth rate

Questions & Answers

Question: How can I tell if data follows a linear, quadratic, or exponential pattern just by looking at the numbers?

Answer: Calculate these values to identify the pattern:

Linear: Calculate first differences (subtract consecutive y-values). If they're constant, it's linear.

  • Example: (1,2), (2,5), (3,8), (4,11) → Differences: 3, 3, 3 → Linear

Quadratic: Calculate first differences, then second differences (differences of differences). If second differences are constant, it's quadratic.

  • Example: (1,1), (2,4), (3,9), (4,16) → First diffs: 3, 5, 7 → Second diffs: 2, 2 → Quadratic

Exponential: Calculate ratios of consecutive y-values (divide). If ratios are constant, it's exponential.

  • Example: (1,2), (2,6), (3,18), (4,54) → Ratios: 3, 3, 3 → Exponential

Always check multiple points to ensure consistency!

Question: When should I use a logarithmic model instead of an exponential model? They seem similar.

Answer: Exponential and logarithmic models are opposites and model very different phenomena:

Exponential models: Represent rapid growth or decay that accelerates over time

  • Population growth
  • Compound interest
  • Radioactive decay
  • Spread of disease

Logarithmic models: Represent growth that starts rapidly but slows over time, approaching a limit

  • Learning curves (improvement slows as skill increases)
  • Diminishing returns
  • Pain perception with increasing medication
  • Perception of loudness with increasing volume

The key difference: Exponential functions continue to accelerate, while logarithmic functions approach a horizontal asymptote.

Question: Can a dataset be modeled by more than one type of function? How do I choose the best one?

Answer: While multiple models might approximate your data, there's usually one that's most appropriate:

Statistical measures: Use correlation coefficient (r) or coefficient of determination (r²). Higher values indicate better fit.

Context matters: Consider what process generated the data. A mathematical fit that doesn't match the real-world situation isn't appropriate.

Residual analysis: Plot the differences between actual and predicted values. Random residuals indicate a good fit.

Domain and range: Ensure the model makes sense over the domain of interest. For example, population can't be negative.

Long-term behavior: Consider how the model behaves outside your data range. Does it make sense for the application?

The "best" model balances mathematical accuracy with real-world applicability.

Question: How do I know if my model is accurate enough? What makes a model "good"?

Answer: A good model meets several criteria:

Accuracy: High correlation coefficient (r close to 1 or -1) and low residuals

Appropriateness: The function type matches the underlying process

Reasonableness: Predictions make sense in the real-world context

Simplicity: The simplest model that adequately explains the data (Occam's Razor)

Quantitative measures include:

  • R² (coefficient of determination): Proportion of variance explained by the model (closer to 1 is better)
  • Root Mean Square Error (RMSE): Average prediction error (lower is better)
  • Residual plots: Should show random scatter, not patterns

Remember: No model is perfect. The goal is to find one that's "good enough" for your purposes.

Question: What if none of the standard models (linear, quadratic, exponential, logarithmic) seem to fit my data well?

Answer: This happens more often than you might think! Here are your options:

Transform the data: Sometimes a transformation makes the data fit a standard model:

  • Take logarithms of both variables for power functions
  • Square the x-values for square root relationships
  • Take reciprocals for inverse relationships

Piecewise functions: Different models for different intervals of data

Higher-order polynomials: Cubic, quartic, etc., but be cautious of overfitting

Trigonometric models: For periodic data

Combination models: Sums or products of basic functions

Non-parametric methods: Techniques that don't assume a specific functional form

If no model fits well, consider whether the relationship is truly predictable or if the data contains too much noise/randomness.