Added several event cullers and SKlearn classifiers

generics/modules/am_sklearn.py

@@ -3,7 +3,10 @@
 from sklearn.svm import LinearSVC, SVC
 from sklearn.neural_network import MLPClassifier
 from sklearn.naive_bayes import MultinomialNB
-from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
+from sklearn.discriminant_analysis import LinearDiscriminantAnalysis, QuadraticDiscriminantAnalysis
+from sklearn.neighbors import NeighborhoodComponentsAnalysis
+from sklearn.linear_model import SGDClassifier, LogisticRegression
+from sklearn.tree import DecisionTreeClassifier
 import numpy as np
 
 class Linear_SVM_sklearn(AnalysisMethod):
@@ -170,3 +173,89 @@ def displayName():
 
 	def displayDescription():
 		return "Multinomial naive bayes implemented in scikit-learn."
+
+class LDA_sklearn(AnalysisMethod):
+
+
+	_NoDistanceFunction_ = True
+
+	def train(self, train, train_data=None, **options):
+		train_data, train_labels = self.get_train_data_and_labels(train, train_data)
+		self._model = QuadraticDiscriminantAnalysis()
+		train_labels = train_labels.flatten()
+		self._model.fit(train_data, train_labels)
+		return
+
+
+	def process(self, docs, Pipe=None, **options):
+		self.train([d for d in docs if d.author != ""], options.get("known_numbers"))
+		test_data = self.get_test_data(docs, options)
+		results = self._model.predict_proba(test_data)
+		results = self.get_results_dict_from_matrix(1-results)
+		return results
+
+	def displayName():
+		return "Linear Discriminant Analysis (sklearn)"
+
+	def displayDescription():
+		return "Linear discriminant analysis implemented in scikit-learn"
+
+class Quadratic_discriminant_analysis(AnalysisMethod):
+
+	_NoDistanceFunction_ = True
+
+	def train(self, train, train_data=None, **options):
+		train_data, train_labels = self.get_train_data_and_labels(train, train_data)
+		self._model = QuadraticDiscriminantAnalysis()
+		train_labels = train_labels.flatten()
+		self._model.fit(train_data, train_labels)
+		return
+
+
+	def process(self, docs, Pipe=None, **options):
+		self.train([d for d in docs if d.author != ""], options.get("known_numbers"))
+		test_data = self.get_test_data(docs, options)
+		results = self._model.predict_proba(test_data)
+		results = self.get_results_dict_from_matrix(1-results)
+		return results
+
+
+	def displayName():
+		return "Quadratic Discriminant Analysis (sklearn)"
+
+	def displayDescription():
+		return "Classifier with quadratic decision boundary. Fits a Gaussian density to each class."
+
+
+class Decision_tree_sklearn(AnalysisMethod):
+
+	_NoDistanceFunction_ = True
+
+	criterion = "gini"
+	splitter = "best"
+	_variable_options = {"criterion":
+		{"options": ["gini", "entropy", "log_loss"], "default": 0},
+        "splitter": {"options": ["best", "random"], "default": 0} #The other options seem like too much for the average user
+	}
+
+	def train(self, train, train_data=None, **options):
+		train_data, train_labels = self.get_train_data_and_labels(train, train_data)
+		self._model = DecisionTreeClassifier(criterion = self.criterion, splitter= self.splitter)
+		train_labels = train_labels.flatten()
+		self._model.fit(train_data, train_labels)
+		return
+
+
+	def process(self, docs, Pipe=None, **options):
+		self.train([d for d in docs if d.author != ""], options.get("known_numbers"))
+		test_data = self.get_test_data(docs, options)
+		results = self._model.predict_proba(test_data)
+		results = self.get_results_dict_from_matrix(1-results)
+		return results
+
+
+	def displayName():
+		return "Decision tree classifier (sklearn)"
+
+	def displayDescription():
+		return "Decision tree classifier implemented in scikit-learn."