dsc 441 asg 3.docx - 1 I Growing Method CRT Validation...

This preview shows page 1 - 6 out of 15 pages.

1) I.
Growing Method : CRT Validation : Holdout Partitioning with a break of 66% preparation and 34% research Maximum tree depth : 10 Minimum cases for parent node : 25 Minimum cases for child node : 12 Impurity measure : Gini Index
II. The final tree has 5 nodes with 3 terminal nodes III. Elements 10, 11 and 1. are the three most notable Lupus data features. Compared to the other factors, they have the greatest relevance. IV. The complexity of a tree decreases after increasing the minimum number of cases for the parent node from 25 to 30 and the child node from 12 to 15. While the accuracy of the testing set reduced, the accuracy of the training set improved. It is possible to justify the reduction in complexity by increasing the minimum number of cases in the child node, which may result in a child not having enough cases so that the parent will not be further separated. 2 1)
In the red wine dataset, there are 6 grades. The skewness, which suggests a minor right skew, is recorded as 0.218. The data is fairly widely spread between the groups, according to the histogram.

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture