Category: s17-busad360

Lesson 15: Calculating Sum of Squares in Python

Thu, Mar 30

Review:

  • Big Data Analytics
  • Intro to Python
  • Hello World Program

Presentation:

Assignment:

  • Write a program to calculate SSxx, SSyy and SSxy using the data set below
    • Data set 1: {(3,5), (5,8), (8,11), (12,10)}
  • Challenge:
    • Write a Python program that does the following
      • Calculates Pearson Correlation Coefficient
      • Calculates Slope and Y-Intercept
      • Prints results to screen
More

Lesson 13: Polynomial Regression

Thu, Mar 16

Review:

  • Measuring Forecast Error

Presentation:

  • Polynomial Regression
    • a form of linear regression
    • only 1 independent variable (x) and 1 dependent variable (y)
    • relationship between x and y is modelled as an nth degree polynomial.
    • 2nd degree polynomial
      • y-hat = b2*x² + b1*x + b0
    • 3rd degree polynomial
      • y-hat = b3*x³ + b2*x² + b1*x + b0
    • Etc
  • Demonstrate Polynomial Regression in Sheets
    • Real Estate Modeling Data
    • Time Series Forecasting
      • Sample data: Crude Oil Prices
      • Re-code Date to Time Period (10/2014 = 1, 11/2014 = 2, and so on)
      • Select “Time Period” and “Price” Columns
        • Insert > Chart
        • Select Scatter Plot as chart type
        • Scroll to bottom, select Trendline > Polynomial
        • Experiment with 2nd, 3rd, 4th, etc degree polynomial
        • Complete activity below with 2nd degree polynomial

Activity:

  • Calculate Price Forecasts
    • Time Periods 12,13,14,15
    • Use Polynomial Regression
    • Use Crude Oil data
  • Measure Forecast Error
    • Calculate MAD
    • See “Actual” tab in Crude Oil Prices data
    • Use actual Oil Prices listed to compare with forecasts

Assignment:

  • Have a good Spring Break!
More

Lesson 12: Measuring Forecast Error

Tue, Mar 14

Review:

Presentation:

  • Forecast Error
    • MAD – mean absolute deviation
      • MAD = ∑|Actual – Forecast|/(Number of Forecasts)
    • MSE – mean square error
      • MSE = ∑(Actual – Forecast)^2/(Number of Forecasts)
    • MAD easier to interpret than MSE
  • Example
    • Actual vs Forecast Selling Prices
    • Actual Forecast |Actual-Forecast| (Actual-Forecast)^2
      55 55 0 0
      27 50 23 527
      217 184 33 1,097
      153 149 4 19
      180 176 4 12
      145 123 22 477
      105 56 49 2,386
      210 225 15 225
      49 130 81 6,547
      83 104 21 461
      251 272 21 428
      145 166 21 460
      294 12,639
      24.5 1,053.3
      MAD MSE

Activity:

  • Calculate MAD for your Exam 2 regression model forecasts vs actual Sell Price.
  • Use a separate piece of paper and submit in-class today.

Assignment:

  • Calculate MAD and MSE for your Exam 2 regression model forecasts vs actual Sell Price
  • Use Google Sheets
  • Submit print copy in-class Thu, Mar 16

 

More

Exam 2 Review

Tue, Mar 7

Review:

Presentation:

  • Exam 2 Review Topics
    • Scatterplots
    • Simple Regression
    • t-test of slope
    • F-test of overall model
    • Multiple Regression
    • Regression Output Interpretation
    • Multicollinearity
    • Binary Variables

Assignment:

  • Study for Exam 2

 

More

Exam 2 Instructions

Instructions:

You will build a multiple regression model to estimate the selling price of single family housing in Pueblo.

You will download and use these 2016 Pueblo real estate sales data to construct the model.

The materials described below will be due during the in-class portion of Exam 2 on [highlight color=”options: yellow, black”]Thu, Mar 9[/highlight].

What to submit:

  1. Prepare a regression output page for your Final Model with footnotes describing any interaction or binary variable use and explaining any other elements of the model that may not be intuitive.
  2. Prepare a 2 page written report describing your regression modeling process and findings. How did you approach this modeling exercise? What variables did you try? Which variables worked best? How did you decide on your final model?

Grading criteria:

  • Accuracy – how good are your sales price estimates when used to estimate 2017 sales
  • Creativity – how creative is your model in terms of identifying useful interactive and binary variables
  • Communication – how well do you describe your model and your modeling process in the report
  • Understanding – how well do you understand your own regression model and regression modeling in general

 

 

 

 

More

Lesson 11: Binary Variables

Tue, Feb 28

Review:

Presentation:

  • Binary variables
    • also known as “Dummy” variables
    • also known as “Indicator” variables
    • value must be either “1” or “0”
  • Demonstration with Real Estate Sample data 
    • create binary variable in separate column with only 1 or 0
    • run regression with the new binary column
    • the coefficient multiplies by 1 if the condition is present
    • the coefficient multiplies by 0 if the condition is not present

Activity:

  • Use the East Pueblo RE Sales 2014-2016 (Sheets)
    • Build upon your multiple regression model from last week
    • Add a Binary variable using the “Sub Area” variable
More

Lesson 10: Multicollinearity

Thu, Feb 23

Review:

  • Multiple Regression

Presentation:

  • Issues in Multiple Regression
    • Multicollinearity
      • multiple variables
      • correlation among independent (x) variables
      • and, correlation with dependent (y) variable
      • correlation is consistently positive or negative
    • Interaction Variables
    • Demonstrate using the Pueblo Sample RE Sales Data
      • Create a new multiple regression model with an interaction variable
      • Mitigate multicollinearity

Activity:

More

Lesson 9: Intro to Multiple Regression

Tue, Feb 21

Review:

  • t-test of slope
  • F-test of overall model

Presentation:

Activity:

  • Complete Problems 1 and 2 using Sheets data
    • Highlight R-squared, SSE and SSyy, p-value for t-test and Significance of F
    • Write the linear equation for y-hat by hand
  • Run Multiple Regression using East Pueblo RE Sales 2014-2016
    • Highlight R-squared, SSE and SSyy, t-test with p-value and F-test with Significance of F
    • Write the linear equation for y-hat by hand
  • Print and submit
More