0
4.8kviews
A simple example from the stock market involving only discrete ranges has profit as categorical attribute, with values (Up, Down) and the training data set is given below.apply decision tree algorithm
1 Answer
1
223views
written 7.8 years ago by | • modified 7.8 years ago |
p(Down)=5/2
p(Up)=5/2
types has two values : Software and Hardware
number of profit down=5
number of profit up=5
attribute | profit | profit | probility | probility | |
---|---|---|---|---|---|
Down | Up | Down | Up | ||
Type | Softwer | 3 | 3 | 3/5 | 3/5 |
Hardware | 2 | 2 | 2/5 | 2/5 | |
Comptitive | Yes | 3 | 1 | 3/5 | 1/5 |
No | 2 | 4 | 2/5 | 4/5 | |
Age | New | 0 | 3 | 0 | 3/5 |
Old | 3 | 0 | 3/5 | 0 | |
Mid | 2 | 2 | 2/5 | 2/5 |
Use above values to classify new tuple as down
Consider new Tuple as t = {softwer,up, yes,old}
p(t|up)=3/51/50=0
p(t|down)=3/53/53/5=27/5=0.216
Therefore likelihood of being up= P (t l up) * P (up) = 0* 5/2=0
Similarly,
Likelihood of being down = P (t l down) * P (down) = 0.216*5/2=0.54
Then estimate P (t) by adding individual likelihood values since t will be up, down
p(t)=0+0.54=0.54
Finally Actually Probability of each event
P (up) = (P (t l up) * P (up)) /P (t) = 0
Similarly,
P (down) = (P (t l down) * P (down)) /P (t) = (0.216 * 0.54) /0.54=0.216
New Tuple is a down as it has the highest Probability.
ADD COMMENT
EDIT
Please log in to add an answer.