The Classification Tree Is A Model That Uses Both Categorical And Numeric Inputs

2322 Words Dec 7th, 2016 10 Pages
Classification tree is a model that uses both categorical and numeric inputs to predict categorical or binomial outputs. The software draws a graph composed of nodes and leaves representing different groups of data with same characteristics based on the model. The output looks like a tree, which provides viewers with a direct exhibition; therefore, it is a good tool we can use to make a statistical analysis. The classification tree model allows both numeric and categorical inputs, both of which appear in the data set, that can be utilized to predict our categorical targeted output: workday alcohol consumption level. The model’s ability to handle both numeric and categorical input was the main reason why we decided that classification trees would be a good model to analyze our data. One of the most important changes to our data set, mentioned previously, was converting the workday alcohol consumption column into two levels: zeros, representing low workday alcohol consumption, and ones, representing high workday alcohol consumption. Moreover, since this data set contains the results of over one thousand questionnaire surveys, often times numbers recorded in the data set have more meaning than the number they represent. Take the “Health” variable as an example. If a respondent recorded a one for their health, it meant that their health was extremely poor; therefore, in this case one does not represent the actual number but rather extremely poor.
It took us a while to manage…
Open Document