add classical methods

traditional_methods/DecisionTree.py

@@ -0,0 +1,20 @@
+from sklearn.tree import DecisionTreeClassifier
+
+import sys
+sys.path.append('../')
+
+from data_loader import MnistLoader
+
+def DT():
+    loader = MnistLoader(flatten=True, data_path='../data', var_per=None)
+    model = DecisionTreeClassifier(criterion='gini')
+
+    model.fit(loader.data_train, loader.label_train)
+    print('model trained')
+    res = model.score(loader.data_test, loader.label_test)
+    print(res)
+
+    return res
+
+if __name__ == '__main__':
+    DT()

traditional_methods/KNN.py

@@ -0,0 +1,20 @@
+from sklearn.neighbors import KNeighborsClassifier
+
+import sys
+sys.path.append('../')
+
+from data_loader import MnistLoader
+
+def KNN():
+    loader = MnistLoader(flatten=True, data_path='../data', var_per=None)
+    model = KNeighborsClassifier(n_neighbors=5)
+
+    model.fit(loader.data_train, loader.label_train)
+    print('model trained')
+    res = model.score(loader.data_test, loader.label_test)
+    print(res)
+
+    return res
+
+if __name__ == '__main__':
+    KNN()

traditional_methods/LR.py

@@ -0,0 +1,20 @@
+from sklearn.linear_model import LogisticRegression
+
+import sys
+sys.path.append('../')
+
+from data_loader import MnistLoader
+
+def LR():
+    loader = MnistLoader(flatten=True, data_path='../data', var_per=None)
+    model = LogisticRegression(penalty='l2')
+
+    model.fit(loader.data_train, loader.label_train)
+    print('model trained')
+    res = model.score(loader.data_test, loader.label_test)
+    print(res)
+
+    return res
+
+if __name__ == '__main__':
+    LR()

traditional_methods/MLP.py

@@ -0,0 +1,20 @@
+from sklearn.neural_network import MLPClassifier
+
+import sys
+sys.path.append('../')
+
+from data_loader import MnistLoader
+
+def MLP():
+    loader = MnistLoader(flatten=True, data_path='../data', var_per=None)
+    model = MLPClassifier(hidden_layer_sizes=(500, 500, 500))
+
+    model.fit(loader.data_train, loader.label_train)
+    print('model trained')
+    res = model.score(loader.data_test, loader.label_test)
+    print(res)
+
+    return res
+
+if __name__ == '__main__':
+    MLP()

traditional_methods/SVM.py

@@ -0,0 +1,20 @@
+from sklearn.svm import SVC
+
+import sys
+sys.path.append('../')
+
+from data_loader import MnistLoader
+
+def SVM():
+    loader = MnistLoader(flatten=True, data_path='../data', var_per=None)
+    model = SVC(kernel='rbf')
+
+    model.fit(loader.data_train, loader.label_train)
+    print('model trained')
+    res = model.score(loader.data_test, loader.label_test)
+    print(res)
+
+    return res
+
+if __name__ == '__main__':
+    SVM()