6.3. Artificial Prevalence protocol with selected usage#

Other way of using the APP protocol is the selected way, where you can choose a specific set of quantifiers to be evaluated, including the ones implemented by the user, see Building a Quantifier for more details. This is useful when you want to evaluate the performance of your own quantifier or a specific set of quantifiers, and compare their results.

To use this approach, you must pass a list of instantiated quantifiers to the models argument. The quantifiers must be any class that inherits from the Quantifier class, linha AggregativeQuantifier or NonAggregativeQuantifier classes, e.g. [CC(), DyS(), EMQ(), myQuantifier()], each one with its parameters setted.

The usage is similar to the general usage, but when using the selected way, you don’t need to pass the learner arument:

from mlquantify.evaluation.protocol import APP
from mlquantify.methods import CC, DyS, EMQ
from sklearn.ensemble import RandomForestClassifier
import numpy as np

X_train = np.random.rand(1000, 20)  # training data
y_train = np.random.randint(0, 2, size=1000)  # training labels
X_test = np.random.rand(1000, 20)  # test data
y_test = np.random.randint(0, 2, size=1000)  # test labels

# list of quantifiers
quantifiers = [CC(RandomForestClassifier()), DyS(RandomForestClassifier()), EMQ(RandomForestClassifier())]  # or 'all', 'aggregative', 'non-aggregative'

app = APP(models=quantifiers,
          batch_size=100,
          n_prevs=20,
          measures=["mae"]
          return_type="table",
          verbose=True)

app.fit(X_train, y_train)
table = app.predict(X_test, y_test)

print(table)

See Protocol for more details on all protocol parameters.

This Page