You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
[V2, 0.0016]
|- 0 ≤ x < 49.5
| [V1, 0.0000]
| |- 1 ≤ x < 1.01
| | +(L, 1)-----------------------------<<<< this supposed to be "s"
|- 49.5 ≤ x < 99.01
| +(L, 1)
[Test]
public void ValueObject_Test_With_Yield_Enumerator()
{
var data = ValueObject.GetData();
var generator = new DecisionTreeGenerator()
{
Descriptor = Descriptor.Create<ValueObject>()
};
var decisionTree = new DecisionTreeGenerator();
var model = generator.Generate(data);
var o = new ValueObject() { V1 = 1, V2 =10 };
var os = model.Predict<ValueObject>(o).R;
Assert.AreEqual("l".Sanitize(), os);
}
The text was updated successfully, but these errors were encountered:
I will take a look at this soon. Thanks for the heads-up.
---UPDATE
In this case the test reflected V2's as 0 - 999. If that was the case, then the answer was indeed correctly generated as the split values when computing information gain in the continuous case does not take into account the data but simply uses the maximum likelihood estimates for splitting (i.e. x < mean <= x). I should add more smarts to this part.
the analyzing was correct and acureate , 0 to 50 = s (small ) , and 0 to 100 "l" large , so the program guessed the equation , but for some reason it put in the both places "l"
and I found in other scenarios some thing , in other example always was taking "very-good" label which was the highest one
The text was updated successfully, but these errors were encountered: