Lesson 4: Analysis of Forecast Error and Curve Fitting

February 1, 2024

Review:

• Linear Regression
• Snowpack and streamflow assignment

Presentation:

• Review streamflow forecasts
• Snowpack n Runoff data
• Plot forecast vs actual
• Measurement of forecast error
  • MAD – mean absolute deviation
    • MAD = ∑|Actual – Forecast|/(Number of Forecasts)
  • RMSE – root mean square error
    • RMSE = √(∑(Actual – Forecast)^2/(Number of Forecasts))

• What can be done if the bivariate relationship is nonlinear?
• Curve fitting
• Use Trendline function in scatter plot
  • Linear
  • Polynomial
• Demonstrate
  • Oil Prices

Activity:

• Use daily 10-year Treasury Rates
• Create a line plot to illustrate recent interest rate activity
  • Use the most recent 90 days
  • Fit a 2nd degree polynomial
  • Capture the polynomial equation
  • Repeat with the most recent 60 days
  • Repeat with the most recent 30 days

Assignment:

• Use the 2nd degree polynomial to forecast oil prices and interest rates
• Decide on the length of the time series for your models