Category: f20-busad360

Lesson 8: Multiple Regression Variable Selection Procedures

Review:

Presentation:

  • Variable Selection Procedures
  • Modeling Objective:
    • Achieve highest R^2 possible
      • For some data this may be below 0.50
      • Don’t sacrifice too much significance for small gains in R^2
    • Achieve high F-statistic
      • Fewer variables = generally preferable (keep it simple)
      • Significance of F (p-value) should be <0.05, lower is better
      • Individual variables should have p-value < 0.05
      • Occasionally it’s appropriate to make exceptions

Assignment:

  • Produce a model to estimate MLB winning percentage
    • Produce a Final Model, aiming to maximize R^2 and F-Statistic
    • Add this model to the Lesson 7 spreadsheet
    • Name the worksheet tab “L8 MLB2013”
  • Assignment 4 Submission
More

Lesson 7: Intro to Multiple Regression

Review:

  • Measuring forecast error with MAD and RMSE
  • Seating chart + Attendance

Presentation:

  • Intro to Multiple Regression
  • Regression Output Interpretation
    • t-statistics
    • p-values
    • R² measures “goodness of fit” (explains % of variance)
    • p-value measures “significance” (probability of result by chance)
    • rank by significance

Activity:

  • Use Microsoft Excel with the Analysis ToolPak Add-in
  • Run Multiple Regression using East Pueblo RE Sales 2014-2016
    • Use “Selg Pr” (selling price) as dependent (y) variable, use “Bd”, “Ba”, “TtlGrSF”, and “DOM” as independent (x) variables
    • Highlight R-squared, coefficients, t-statistics and p-values
    • Type the linear equation for y-hat in a separate cell
    • List, in rank order, the variables from most significant to least significant

Assignment:

  • Save the regression output in Excel
    • Name the worksheet tab “L7 East Pueblo”
    • You will submit this spreadsheet file after Lesson 8 (Thu)
More

Lesson 6: Measuring Forecast Error

Review:

  • Time Series Forecasting

Presentation:

  • Forecast Error
    • MAD – mean absolute deviation
      • MAD = ∑|Actual – Forecast|/(Number of Forecasts)
    • RMSE – root mean square error
      • RMSE = √(∑(Actual – Forecast)^2/(Number of Forecasts))
    • Scatterplot comparison – actual vs forecast
    • Residual analysis
    • Outliers
  • Demonstrate with Oil Prices
  • Compare to Actual Oil Prices

Activity:

Assignment:

  • No assignment due this week.
  • Use the time to get caught up or get ahead in this or other classes.
  • We will begin Multiple Regression next week

More

Lesson 5: Time Series Analysis (via Zoom due to weather)

Zoom Lesson Recording

Review:

  • Attendance
  • Simple Regression
  • Assignment 3

Presentation:

  • Time Series Analysis
    • Scatter plots
    • Line plots
    • Trend lines
      • Linear
      • Polynomial
    • Forecasting
    • Oil Prices

Activity:

  • Use daily 10-year Treasury Rates
  • Create a line plot to illustrate the last 60 days of interest rates
  • Fit a 2nd degree polynomial
  • Use the 2nd degree polynomial to forecast interest rates
  • Generate forecasts for today and tomorrow (Sep 8-9)

Assignment:

  • To be posted before class Thursday
More

Lesson 4: Simple Linear Regression

Review:

  • Attendance
  • Scatterplots
  • Trendline

Presentation:

  • Simple Linear Regression
    • Bivariate (2 variables)
    • i.e., one x and one y variable
  • Demonstrate Estimation with Linear Regression
    • Pueblo RE 2014 – 100 rows
    • Total SqFt = x/independent/explanatory variable
    • Selling Price = y/dependent/response variable
    • Generate Scatterplot
      • Fit Trendline
      • Label with Equation and R²
    • Replicate linear equation in spreadsheet
    • Estimate Price assuming SqFt = 1200, 1600, 2000

Activity:

Assignment:

More

Lesson 3: Visualizing Bivariate Relationships with Scatterplots

Review:

  • Attendance
  • Histograms
  • DataCamp

Presentation:

Activity:

  • Create Scatterplots
    • Use this data set: Pueblo RE 2014 – 100 rows
    • Beds (x) vs Selling Price (y)
    • Baths (independent) vs Selling Price (dependent)
    • Total SqFt (explanatory) vs Selling Price (response)
  • Identify the strongest correlation
    • Beds, Baths or Total SqFt

Assignment:

More

Lesson 2: Visualizing Data Distributions with Histograms

Review:

  • Attendance
  • DataCamp

Presentation:

Visualizing Data Distributions

  • Frequency distribution
    • measure of frequency within a range of values
  • Exam Scores
x̅ = 83.7
x̅ = 86.3
x̅ = 81.2
  • Wildfires
  • Tropical Storms in the Atlantic
Íomhá
NYTimes 8/27/2020
  • Population

How to make a histogram in Sheets

  • Select the variable you want to display (one at a time, for now)
  • Go to Insert > Chart > Set “Chart Type” to “Histogram”
  • Here’s some US County Population Data for illustration

Activity:

  • Make a histogram for Colorado County Populations (n=64)

Assignment:

  • Complete the “Data Analysis in Spreadsheets” Course in DataCamp
More

Lesson 1: Introduction to Advanced Statistics

Presentation:

In-Class Activity:

  • Join my classroom on DataCamp.com
    • Check your csupueblo.edu email; there should be an invitation from DataCamp
    • Click on the link to join my classroom
    • Your account is free – don’t pay for premium access!
    • If you have trouble, email [email protected]
  • Begin Introduction to Spreadsheets

Assignment:

More