Geographically Weighted Regression

Exploratory Regression

Exploratory Regression results with all 8 census variables

Exploratory Regression results with all 8 census variables

Significance of Variables

Significance of Variables

Chosen Explanatory Variables: Education, Immigrants, Visible Minority

Geographically Weighted Regression

After completing an OLS model, a GWR was performed with the dependent variable (Density of Incidences) and the 3 chosen Explanatory variables (Education, Immigrants, and Visible Minority) to  fit a local regression equation for each feature (DA). GWR was completed with an Adaptive kernel type, and an AICc bandwidth. Coefficient raster surfaces were manually made by converting DA polygons to rasters (cell size 50) and displayed in stretched symbology – Standard Deviation.

 

Additional Census Variables

An Exploratory Regression was repeated to include additional census variables: % Labour Force by Industry: Mining, quarrying, and oil and gas extraction, % Labour Force by Occupation: Natural resources, agriculture and related production occupation and % Labour Force by Industry: Agriculture, forestry, fishing and hunting.

Exploratory Regression with Additional Variables

Exploratory Regression with Additional Variables

 

Chosen Explanatory Variables: Education, Immigrants, Mining Industry, Natural Resources and Agriculture

An OLS and GWR were completed once again following the same parameters as specified above.

 

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Spam prevention powered by Akismet