Report on Data Mining and Visualization for Business Intelligence

10 Pages1048 Words94 Views

Added on 2020-04-07

Report on Data Mining and Visualization for Business Intelligence

Added on 2020-04-07

BookmarkRelated Documents

Data Mining and Visualization forBusiness IntelligenceAssignment – 2[Pick the date]STUDENT ID

Report on Data Mining and Visualization for Business Intelligence_1

Question 1(a)PCA Output using XL MinerThe principal component matrix helps in identification of useful features which need to be usedfor comparison purpose. For this, typically, the magnitude of the coefficients in the principalcomponent matrix would be critical while the sign just denotes the direction. For example,consider the principal component 1. The two key features for this principal component would bex2 and x1 considering that these two features provide the highest coefficient magnitude. This canbe then extended to others and a list of all key features for the various principal components canbe obtained. It is essential that these features collectively indicate the critical characteristics forthe utility companies which must be considered for further analysis. Based on the PCA output,the critical features have been identified as shown below.1

Report on Data Mining and Visualization for Business Intelligence_2

Also, it needs to be discussed if normalisation is necessary for the dataset provided. The utility ofnormalisation in PCA arises since the measurement scale of difference variables tends to differwhich leads to higher variance magnitude for those variables that have bigger values.Considering the maximum variance emphasis for the PCA, the results produced could beincorrect. Even though for the US utility data, results are not so skewed so as to warrantnormalisation based on the variance matrix alone. However, considering the scale of certainvariables such as x3, it would be appropriate to conclude that data normalisation would lead tobetter results. (b) The list of advantages and disadvantages of principal component analysis are shown below:List of advantages of principal component analysis:1.Reduces high multi-dimensional set of data into fewer number dimensional set of data. Thisis termed as dimensional reduction. 2.PCA helps to decide the structure of components. 3.Variables would be defined in the form of coordination structure.2

Report on Data Mining and Visualization for Business Intelligence_3

End of preview

Want to access all the pages? Upload your documents or become a member.

Data Mining & Visualization for Business Intelligence

|781

|55

Data Mining and Visualization Assignment

|1040

|50

Data Mining and Visualization | Assessment

|10

|1183

|87

Variance Matrix - A key feature of DMVBI

|1369

|444

Assignment Data Mining & Visualization for Business Intelligence

|1204

|86

The Advantages of PCA Method in Data Mining - Desklib

|11

|1673

|84

Report on Data Mining and Visualization for Business Intelligence