merge

Author: cosmic-cortex

examples/keras_integration.py

@@ -7,7 +7,7 @@
 import numpy as np
 from keras.datasets import mnist
 from keras.models import Sequential
-from keras.layers import Conv2D, MaxPooling2D, Dropout, Flatten, Dense
+from keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPooling2D
 from keras.wrappers.scikit_learn import KerasClassifier
 from modAL.models import ActiveLearner
 
@@ -18,8 +18,10 @@ def create_keras_model():
     This function compiles and returns a Keras model.
     Should be passed to KerasClassifier in the Keras scikit-learn API.
     """
+
     model = Sequential()
     model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=(28, 28, 1)))
+    model.add(Conv2D(64, (3, 3), activation='relu'))
     model.add(MaxPooling2D(pool_size=(2, 2)))
     model.add(Dropout(0.25))
     model.add(Flatten())
@@ -68,20 +70,21 @@ def create_keras_model():
 learner = ActiveLearner(
     estimator=classifier,
     X_training=X_initial, y_training=y_initial,
-    verbose=0
+    verbose=1
 )
 
 # the active learning loop
 n_queries = 10
 for idx in range(n_queries):
     query_idx, query_instance = learner.query(X_pool, n_instances=200, verbose=0)
+    print(query_idx)
     learner.teach(
         X=X_pool[query_idx], y=y_pool[query_idx],
-        verbose=0
+        verbose=1
     )
     # remove queried instance from pool
     X_pool = np.delete(X_pool, query_idx, axis=0)
     y_pool = np.delete(y_pool, query_idx, axis=0)
 
 # the final accuracy score
-print(learner.score(X_test, y_test, verbose=0))
+print(learner.score(X_test, y_test, verbose=1))

modAL/utils/combination.py

@@ -1,14 +1,6 @@
 import numpy as np
 
 
-def check_init(obj, typ):
-    if obj is None:
-        return typ()
-    else:
-        assert isinstance(obj, typ), 'obj must be of type %s' % typ.__name__
-        return obj
-
-
 def make_linear_combination(*functions, weights=None):
     """
     Takes the given functions and makes a function which returns the linear combination
@@ -38,8 +30,8 @@ def make_linear_combination(*functions, weights=None):
                                                'same as the number of given functions'
 
     def linear_combination(*args, **kwargs):
-        return np.sum([weights[i]*functions[i](*args, **kwargs)
-                       for i in range(len(weights))], axis=0)
+        return np.sum((weights[i]*functions[i](*args, **kwargs)
+                       for i in range(len(weights))), axis=0)
 
     return linear_combination
 
@@ -79,15 +71,12 @@ def product_function(*args, **kwargs):
     return product_function
 
 
-def make_query_strategy(utility_measure, selector, utility_kwargs, selector_kwargs):
-
-    utility_kwargs, selector_kwargs = check_init(utility_kwargs, dict), check_init(selector_kwargs, dict)
-
+def make_query_strategy(utility_measure, selector):
     # TODO: check for the signatures of utility_measure and selector
 
-    def query_strategy(classifier, X, utility_kwargs, selector_kwargs):
-        utility = utility_measure(classifier, X, **utility_kwargs)
-        query_idx, query_instance = selector(utility, **selector_kwargs)
+    def query_strategy(classifier, X):
+        utility = utility_measure(classifier, X)
+        query_idx, query_instance = selector(utility)
         return query_idx, query_instance
 
     return query_strategy