Decision Tee Expert

You are to build a decision tree to determine if an individual is suffering from diabetic hypoglycemia, diabetic hyperglycemia, pre-diabetes, food poisoning, shock, or normal (no condition). You must obtain the decision tree tool See5 (requires either Windows or Linux operating system). Follow these steps: Navigate to: [login to view URL] (Links to an external site.) Scroll down a little, to the second set of links and click on the appropriate See5/C5.0 download link for your operating system. Run the installer you just downloaded You are now ready to build Decision Trees Before building the tree and testing it, you must first create a data definition and training files. To do this use the spreadsheet: [login to view URL] Download [login to view URL] See [login to view URL] (Links to an external site.) for directions on how to set up your .names, .data, and .test files. The .names file you will need to create following the examples given in the tutorial, but for the data represented in the spreadsheet. Excel should be able to produce a comma separated data and test files using the SaveAs feature. The diagnosis appears in the last column and other data you can use is in the other columns. It is up to you to decide which columns of data to use, but you must use at least 4. Try this exercise first without using the Blood Sugar values as one of your 4 selected values and then a second time with the Blood Sugar values as one of the 4 or more data items your tree uses. (Keep this file as you will be using it in your other projects as well) Use the spreadsheet to determine which variables you want See5 to consider (defined in your .names file) and create a new sheet with these values (remember, you must have at least 4 tests and the diagnosis). Now separate your selected values into two spreadsheets each with different rows of items and save each sheet as a comma separated text file. You will have to rename the file extensions from .txt to the appropriate file extensions specified in the tutorial. These will be used to train and validate your decision tree. Note: do not save the column titles, you must remove these before saving your data as a text file. Once you have all 3 files created, start the See5 program you installed. Click the Locate Data button near the top on the left side. Navigate to where you have your 3 files stored and select the .data file you created. See5 will automatically find your .names file and your .test file, so long as the file names are the same. Such as [login to view URL], [login to view URL], diagnosis.test.