6. Construct a decision tree a. Build a decision tree for the O-ring data set using information gain ratio to decide the split attribute from table 2. Stop split when number of instance less than 2. b. Post pruning the tree based on train dataset with confidence limit = 20% (z = 0.25). Table 2 O-ring Attribute variables Target variable number of O- Launch Leak-Check Number of O- rings temperature Pressure rings with Stress 1 6 66 50 0 2 6 70 50 1 3 6 69 50 0 4 6 68 50 0 5 6 67 50 0 6 6 72 50 0 7 6 73 100 0 8 6 70 100 0 9 6 57 200 1 10 6 63 200 1 11 12 13 14 15 16 666666 6 70 200 1 78 200 0 67 200 0 53 200 2 67 200 0 75 200 0 17 6 70 200 0 18 6 81 200 0 19 6 76 200 0 20 6 79 200 0 21 6 75 200 0 22 22 6 76 200 0 23 6 58 200 1 Table 4: O-ring with missing value Attribute variables Target variable number of O- Launch Leak-Check Number of O- rings temperature Pressure rings with Stress 1 6 66 50 0 2 6 70 50 1 3 6 69 50 0 4 6 68 50 0 5 6 67 50 0 6 6 72 50 0 7 6 73 100 0 8 6 ? 100 0 9 6 57 200 1 10 6 63 200 1 11 6 70 200 1 12 6 78 200 0 13 6 67 ? 14 6 53 200 15 6 67 200 16 6 75 200 0200 17 6 70 200 0 18 6 81 200 0 19 6 76 200 0 20 6 79 200 0 21 6 75 200 0 22 6 76 200 0 23 6 58 200 1