Spatial Analysis

After all the factor images have been standardized or converted to fuzzy images, the Multi-criteria Evaluation (MCE) begins. The first step in the MCE is to perform a series of pairwise comparison between factors. The purpose of this procedure is to obtain computer generated factor weight for each individual factor image thus make them comparable. Specifically, factor images that are relative more important would be assigned higher weight than those that are less important. In general, sum of all weights from all the factors would equal to 1. Pair wise comparison was done through the Multi-Objective Decision Wizard specifically via AHP (analytical hierarchy process). Pair wise comparison was performed in a matrix style format (See table below).

In general, the numbers from the above table represent the relative importance between factors. For example, when the same pair of factors were compared, the number 1 is assigned, as both are equally important. Conversely, when the factors such as average annual family income was compared to Chinese speaking, a fraction 1/2 was assigned. This means that factor such as average annual family income is 1/2 as important to that of Chinese speaking. All the numbers / fractions assigned to the matrix table are all subjective and based on personal opinion. Subsequently, once the matrix table was completed, it was inputed into the AHP module were the factor weights were generated. See output below:

The eigenvector of weights is :

chinese_spk_fuzzy : 0.4185
family_income_fuzz : 0.2625
4_5_family_fuzzy : 0.1599
2_3_cars_fuzzy : 0.0973
owner_occ_fuzz : 0.0618

Consistency ratio = 0.02
Consistency is acceptable.

From the output above, one can observe that the factor Chinese speaking have the highest weight value, while the owner occupational unit have the lowest. In turn this mean that relatively more weight or emphasis would be given to the Chinese speaking factors during the Multi-Critieria Evaluation process. The consistency ratio (CR) in general, represents the overall quality of the pair wise comparison, specifically whether the comparison between various pair of factor make logical sense. An acceptable consistency ratio needs to be less than 0.1, so in this particular scenario where CR is 0.02 the overall pair wise comparison was acceptable. Below is the generated MCE image resulting from the given factor weights. The dark red shaded regions represent areas of interest in the city of Chicago.

Although several potential areas of interest have been highlighted from the MCE, only areas that have suitability scale of 220 and above were selected through reclass module in Idrisi. After reclass, only one particular area was highlighted as it is shown below. The dimension of this particular area is approximately *59 square kilometres.

* Area value was obtained through area module that was ran in the software Idrisi. Click Here to see cartographic model for the calculated area.

To see the cartographic model for this project click here

Previous Page Next Page

Back to Home Page