Fitting Exponential Models to Data: Fresh Take

  • Build an exponential model from data
  • Build a logarithmic model from data
  • Build a logistic model from data

Build an Exponential Model from Data

The Main Idea

  • Exponential Regression:
    • Used when data grows or decays at an increasing rate
    • Model form: [latex]y = ab^x[/latex]
    • If [latex]b > 1[/latex]: exponential growth
    • If [latex]0 < b < 1[/latex]: exponential decay

Key Techniques

  1. Using a Graphing Calculator:
    • Enter data in lists L1 (x-values) and L2 (y-values)
    • Create a scatter plot to verify exponential pattern
    • Use “ExpReg” from STAT CALC menu
    • Record values for [latex]a[/latex] and [latex]b[/latex]
  2. Using Technology (Desmos):
    • Create a table with your data
    • Enter [latex]y_1 \sim ab^{x_1}[/latex] below the table
    • The tool displays values for [latex]a[/latex] and [latex]b[/latex]
A credit card balance grows each month after graduation. The table shows the debt over 8 months:[latex]\begin{array}{|c|c|c|c|c|c|c|c|c|} \hline \text{Month} & 1 & 2 & 3 & 4 & 5 & 6 & 7 & 8 \\ \hline \text{Debt (\$)} & 620 & 761.88 & 899.80 & 1039.93 & 1270.63 & 1589.04 & 1851.31 & 2154.92 \\ \hline \end{array}[/latex]Use exponential regression to fit a model to these data. If spending continues at this rate, what will the debt be after one year (month 12)?

Build a Logarithmic Model from Data

The Main Idea

  • Logarithmic Regression:
    • Used when data increases or decreases rapidly at first, then slows
    • Model form: [latex]y = a + b\ln(x)[/latex]
    • All input values must be positive
    • If [latex]b > 0[/latex]: increasing model
    • If [latex]b < 0[/latex]: decreasing model

Key Techniques

  1. Using a Graphing Calculator:
    • Enter data in lists L1 and L2
    • Verify scatter plot shows logarithmic pattern
    • Use “LnReg” from STAT CALC menu
    • Record model as [latex]y = a + b\ln(x)[/latex]
Life expectancy has been increasing since 1900. The table shows average life expectancies of Americans:[latex]\begin{array}{|c|c|c|c|c|c|c|} \hline \text{Year} & 1900 & 1910 & 1920 & 1930 & 1940 & 1950 \\ \hline \text{Life Exp.} & 47.3 & 50.0 & 54.1 & 59.7 & 62.9 & 68.2 \\ \hline \end{array}[/latex][latex]\begin{array}{|c|c|c|c|c|c|c|} \hline \text{Year} & 1960 & 1970 & 1980 & 1990 & 2000 & 2010 \\ \hline \text{Life Exp.} & 69.7 & 70.8 & 73.7 & 75.4 & 76.8 & 78.7 \\ \hline \end{array}[/latex]Let [latex]x = 1[/latex] for 1900, [latex]x = 2[/latex] for 1910, etc. Use logarithmic regression to predict life expectancy in 2030.

Build a Logistic Model from Data

The Main Idea

  • Logistic Regression:
    • Used when growth starts rapidly then levels off to an upper limit
    • Model form: [latex]y = \frac{c}{1 + ae^{-bx}}[/latex]
    • [latex]c[/latex] is the limiting value (carrying capacity)
    • Initial value is [latex]\frac{c}{1+a}[/latex]

Key Techniques

  1. Using a Graphing Calculator:
    • Enter data in lists L1 and L2
    • Verify scatter plot shows logistic pattern (S-shaped curve)
    • Use “Logistic” from STAT CALC menu
    • Record values for [latex]a[/latex], [latex]b[/latex], and [latex]c[/latex]
Harbor seal populations in the Wadden Sea from 1997-2012:[latex]\begin{array}{|c|c|c|c|c|} \hline \text{Year} & 1997 & 1998 & 1999 & 2000 \\ \hline \text{Population (thousands)} & 3.493 & 5.282 & 6.357 & 9.201 \\ \hline \end{array}[/latex](data continues through 2012 with population reaching 25.108)Let [latex]x = 0[/latex] for 1997. Use logistic regression to:
a. Fit a model to the data
b. Predict seal population in 2020
c. Find the carrying capacity