CODEWITHSUNDEEP
pythonscikit-learn
sergzach
sergzach

Reputation: 6764

Can DecisionTreeClassifier be used with both binary and multiclass labels simultaneously?

Can I fit the classifier with binary and multiclass labels to predict a result?

Multiclass labels can have more then 2 values, binary labels only can have 2.

Example (the first parameter in X is multiclass, the second is binary - [-1,1]):

from sklearn import tree
X = [[0, -1], [2, 1], [1, -1] ]
Y = [0, 1, 1]
clf = tree.DecisionTreeClassifier()
clf = clf.fit(X, Y)
clf.predict([[1, 1]])

Upvotes: 0

Views: 320

Answers (1)

Ando Saabas
Ando Saabas

Reputation: 1977

I think you are confusing labels (usually represented with Y variable as you also do in your example) with features (the matrix X in your example). When talking about binary or multiclass labels, it's the output/response variable Y that is usually meant.

The input variables (values in matrix X) can be arbitrary integers or floats, so your example is perfectly valid.

Upvotes: 2

Related Questions