def knn_classify(df_gender, predicted_variable):
X_train, X_test, y_train, y_test = Utils.split_data(df_gender)
neigh = KNeighborsClassifier(n_neighbors=18)
neigh.fit(X_train, y_train)
pickle.dump(neigh, open("resources/KNNimages_" + predicted_variable + ".sav", 'wb'))
y_pred = neigh.predict(X_test)
print("KNN acc: ", accuracy_score(y_test, y_pred))
def kernel_estimation(df_gender):
X_train, X_test, y_train, y_test = Utils.split_data(df_gender)
rbf_feature = RBFSampler()
X_features = rbf_feature.fit_transform(X_train)
clf = SGDClassifier()
clf.fit(X_features,y_train)
print("Kernel Density acc: ", clf.score(X_features, y_train))
def logistic_regression_customized(df):
X_train, X_test, y_train, y_test = Utils.split_data(df)
clf = LogisticRegression()
clf.fit(X_train, y_train)
y_pred = clf.predict_proba(X_train)
y_df = pd.DataFrame(y_pred)
chos = y_df.max(axis=1)
chos_df = pd.DataFrame(chos)
indexes = set()
for index, row in chos_df.iterrows():
if row[0] < 0.30:
indexes.add(index)
y_prediction = clf.predict(X_test)
for j in range(y_prediction.shape[0]):
if j in indexes:
y_prediction[j] = "xx-24"
print("logistic regression acc: ",
accuracy_score(y_test, y_prediction))
def svm_estimation(df_gender):
X_train, X_test, y_train, y_test = Utils.split_data(df_gender)
clf = svm.SVC(gamma='scale')
clf.fit(X_train, y_train)
y_pred = clf.predict(X_test)
print("SVM acc: ", accuracy_score(y_test, y_pred))
def sgd_classify(df_gender):
X_train, X_test, y_train, y_test = Utils.split_data(df_gender)
clf = SGDClassifier(loss="hinge", penalty="l2")
clf.fit(X_train, y_train)
y_pred = clf.predict(X_test)
print("sgd acc: ", accuracy_score(y_test, y_pred))
