Need to analyize decision tree that shown down for each trace arrow: Samples/Tut

Question

Need to analyize decision tree that shown down for each trace arrow:

Samples/Tutorials/Basics/08/process RapidMiner Studio Educational 7.4.000 a AGUDELL2HZB File Edit Process View Connections Cloud Settings Extensions Design Views: Results Result History X Tree (Decision Tree Zoom Sex Female Male Graph Mode No of Siblings or Spouses on Board 4.500 4.500 No Description Tree No of Parents or Children on Board Node Labels s 5.500 5.500 Edge Labels No Passenger Class Annotations Third First Second Yes Age Passenger Fare 32 is 32.881 56 s 56 No Questions?

Explanation / Answer

Decision tree is one of the predictive data mining technique.decision tree have a root node,childern nodes(non leaf nodes)and leaf nodes.each node in tree is a test of attribute.child nodes are outcome of attribute.

leaf node repregents class labels,there are two class labels :yes class ,no class.

step 1: Root Node is sex.root is test of attribute.

the outcome is two values Female or male.(each attribute have values sex is binary attribute it have two values they are female and male)

if sex=male then assign the class label 'No'.otherwise go for test of attribute.

if sex= female then go for test of attribute.(next attribute)

step 2: test of attribute is no of siblings or spouses it is continuous values attribute it has several vales gruop them into binary way(A>value or A< Value ) or multiway.   

if no of siblings or spouses are greater than 4.500 then assign class label='No'.

if no of siblings or spouses less than orequals to 4.500 then go for next test of attribute.

step 3:the test of attribute is no of parents or childern on board it also continuous valued attribute it have many values splitting into two ways(attribute value>5.5 or attribute value<=5.5).

if no of parents or childern on board is greater than 5.500 then assign class label='No'.

if no of parents or childern on board is less than or equal to 5.500 then go for next test of attribute to predict the class label.

step 4:the intermediate node test of attribute is passenger class.it is ordinal attribute the values must follow particular order.here the order first, second, third

the passanger attribute has three oucomes.three childerns created.

if the passenger class is first then assign class label='Yes'.

if the outcome is second then go for test of attribute age.

if outcome is third then go for next test of attribute contion passenger fare.

step5: test of attribute age.

the age attribute has three outcomes

if age greater than 56 assign class label='No'.

if age less than or equal to 56 assign class label='yes'.

if there is no age value available of an attribute simply assign class label 'yes'

step6:Test of attribute

if passenger class is outcome is 'third' then test of attribute is passenger fare it is also continuous values attribute split the values bainary way(passenger fare>32 or passenger fare<=32.881)

if passenger fare greater than 32 assign class label 'No'

if passenger fare less than or eual to then asign class label 'yes'.

Note:Leaf nodes repregents class labels they are 'yes' class or 'No' class.

Related Questions

Navigate

• Browse (All)
• Browse N
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.