Tutorial on Machine Learning

  1. The input to an ID3 algorithm is shown in the table below. Describe how the information gain calculations for the first split in the decision tree are done if the split is based on the texture attribute:

    The attributes and their possible values are:
    texture: smooth wavy rough
    temperature: cold cool warm hot
    size: small medium large

    texturetemperaturesizeclass
    smoothcoldlargeyes
    smoothcoldsmallno
    smoothcoollargeyes
    smoothcoolsmallyes
    smoothhotsmallyes
    wavycoldmediumno
    wavyhotlargeyes
    roughcoldlargeno
    roughcoollargeyes
    roughhotsmallno
    roughwarmmediumyes

  2. Given the subtree of a decision tree shown below, and using the Laplace error estimate and the criterion defined in lectures, should the the children of the subtree be pruned or not?

                        ( [8, 5] )
                         /      \
                        /        \
                       /          \
                    [6,2]        [4,1]
    

Solutions when available

CRICOS Provider Code No. 00098G