Y_train = np.array(Y_train)
X_test = np.array(X_test)
Y_test = np.array(Y_test)
# Hyperparameters
epochs = 0
lRate = .01
C = ['100/873', '500/873', '700/873']
epochs = 100
print('Running Stochastic Gradient Descent for SVM...\n')
#while(epochs < maxIter):
lRate = .01
for c in C:
w = SVM.SGD(X_train, Y_train, eval(c), lRate, epochs)
# Get predictions from model
Y_predict = []
correct = 0
n = len(X_train)
for i in range(n):
yp = np.sign(np.dot(w, X_train[i]))
Y_predict.append(yp)
# Calculate accuracy
for i in range(len(Y_train)):
if Y_train[i] == Y_predict[i]:
correct += 1
accuracy = correct / n
print('C = ', c)
print('Learned weight vector: ', w)
print('Accuracy: ', accuracy, '\n')
