def prob_3(weighted_d = False):
test_arff = Arff("housing_testing_data.arff")
train_arff = Arff("housing_training_data.arff")
test_arff.shuffle()
train_arff.shuffle()
test_arff.normalize()
train_arff.normalize()
K = [1, 3, 5, 7, 9, 11, 13, 15]
A = []
for k_hat in K:
test_data = np.hstack((test_arff.get_features().data, test_arff.get_labels().data))
train_data = np.hstack((train_arff.get_features().data, train_arff.get_labels().data))
KNNC = KNNClassifier(k_hat, train_data, test_data)
A.append(KNNC.get_accuracy_regress(weighted_d))
plt.plot(K, A, label="")
t = "KNN Regression M.S.E Housing"
if weighted_d:
t += "(weighted-d)"
weighted_d
plt.title(t)
plt.xlabel("K")
plt.ylabel("M.S.E")
# plt.legend()
plt.show()
def prob4h():
arff = Arff('datasets/abalone.arff', label_count=0)
arff.normalize()
domain = np.arange(2, 8)
print('single link --------------------')
hoc = HAC()
hoc.train(arff, printk=domain, silhouette=True)
print('complete link -----------------------')
hoc = HAC(simple=False)
hoc.train(arff, printk=domain, silhouette=True)
def prob4():
arff = Arff('datasets/abalone.arff', label_count=0)
arff.normalize()
domain = np.arange(2, 8)
ssekmm = []
for k in domain:
km = KMeans(k)
ssek = km.train(arff)
ssekmm.append(ssek)
print(km.calc_silhouette_score())
def test_cases():
# test_1()
attr_types = [
"real",
"real",
"real",
"real",
"cat",
"real",
"cat",
"real",
"real",
"cat",
"real",
"cat",
"cat",
"cat",
"cat",
"cat",
"cat"
]
attr_idx = [
[],
[],
[],
[],
['none','tcf','tc'],
[],
['none','ret_allw','empl_contr'],
[],
[],
['yes','no'],
[],
['below_average','average','generous'],
['yes','no'],
['none','half','full'],
['yes','no'],
['none','half','full'],
['bad','good']
]
k = 5
arff = Arff("labor.arff")
arff.normalize()
features = arff.get_features().data
labels = arff.get_labels().data
# attributes = arff.get_attr_names()
data = np.hstack((features, labels))[:, 1:]
kmc = KMC(k, data, data, attr_types, attr_idx)
kmc.train(tol=0)
def prob_5():
cont_mask = [1, 2, 7, 10, 13, 14, 16]
cate_mask = [0, 3, 4, 5, 6, 8, 9, 11, 12, 15]
arff = Arff("credit_approval_data.arff")
arff.shuffle()
arff.normalize()
n = len(arff.get_labels().data)
t = int(n * .7)
train_data = arff.create_subset_arff(row_idx=slice(0, t, 1))
test_data = arff.create_subset_arff(row_idx=slice(t, n, 1))
test_data = np.hstack((test_data.get_features().data, test_data.get_labels().data))
train_data = np.hstack((train_data.get_features().data, train_data.get_labels().data))
#b,30.83,0,u,g,w,v,1.25,t,t,01,f,g,00202,0,+
dist_matrix = np.ones((16, 16))
np.fill_diagonal(dist_matrix, 0)
KNNC = KNNClassifier(8, train_data, test_data)
print(KNNC.get_accuracy_mixed(cate_mask, cont_mask, dist_matrix))
def prob_2(weighted_d = False):
""" """
k = 3
test_arff = Arff("magic_telescope_testing_data.arff")
train_arff = Arff("magic_telescope_training_data.arff")
test_arff.shuffle()
train_arff.shuffle()
# attributes = test_arff.get_attr_names()
test_data = np.hstack((test_arff.get_features().data, test_arff.get_labels().data))
train_data = np.hstack((train_arff.get_features().data, train_arff.get_labels().data))
KNNC = KNNClassifier(k, train_data, test_data)
acc = KNNC.get_accuracy(weighted_d)
test_arff.normalize()
train_arff.normalize()
n_test_data = np.hstack((test_arff.get_features().data, test_arff.get_labels().data))
n_train_data = np.hstack((train_arff.get_features().data, train_arff.get_labels().data))
n_KNNC = KNNClassifier(k, n_test_data, n_train_data)
acc_n = n_KNNC.get_accuracy(weighted_d)
# print(np.array([[acc,acc_n]]))
print(acc,acc_n)
# show_table(["Not Normalized" "Normailzed"], ["Accuracy"], np.array([[acc,acc_n]]), title = "Normalized vs Non-normalized, k=3")
K = [1, 3, 5, 7, 9, 11, 13, 15]
A = []
for k_hat in K:
# n_test_data = np.hstack((test_arff.get_features().data, test_arff.get_labels().data))
# n_train_data = np.hstack((train_arff.get_features().data, train_arff.get_labels().data))
n_KNNC = KNNClassifier(k_hat, n_train_data, n_test_data)
A.append(n_KNNC.get_accuracy(weighted_d))
plt.plot(K, A, label="")
t = "KNN Accuracy Telesc. "
if weighted_d:
t += "(weighted-d)"
plt.title(t)
plt.xlabel("K")
plt.ylabel("Accuracy")
# plt.legend()
plt.show()
def prob_6():
""" """
k = 3
test_arff = Arff("magic_telescope_testing_data.arff")
train_arff = Arff("magic_telescope_training_data.arff")
test_arff.shuffle()
train_arff.shuffle()
test_arff.normalize()
train_arff.normalize()
K = [1, 3, 5]
T = []
A = []
T_KSM = []
A_KSM = []
for k_hat in K:
test_data = np.hstack((test_arff.get_features().data, test_arff.get_labels().data))
train_data = np.hstack((train_arff.get_features().data, train_arff.get_labels().data))
KNNC = KNNClassifier(k_hat, train_data, test_data)
t = time.time()
A.append(KNNC.get_accuracy())
T.append(time.time() - t)
KNNC.induce_KSM()
t = time.time()
A_KSM.append(KNNC.get_accuracy())
T_KSM.append(time.time() - t)
ax = plt.axes(projection='3d')
ax.plot(K, A, T, label="No-KSM")
ax.plot(K, A_KSM, T_KSM, label="KSM")
ax.set_xlabel('K')
ax.set_ylabel('Accuracy')
ax.set_zlabel('Time')
t = "KNN Accuracy w/ IKSM"
plt.title(t)
plt.legend()
plt.show()
def prob_3():
# Use regression knn on housing price prediction dataset
train = Arff('datasets/housing_train.arff')
test = Arff('datasets/housing_test.arff')
train.normalize()
test.normalize()
krange = np.arange(1, 16, 2)
mses = []
for k in krange:
knn = KNN(k)
preds = knn.knn(train.get_features(), train.get_labels(),
test.get_features())
mse = sum((preds - np.ravel(test.get_labels().data))**2) / len(preds)
mses.append(mse)
plt.plot(krange, mses)
plt.title("K Size Versus MSE on Housing Prices")
plt.xlabel("K")
plt.ylabel("Mean Squared Error")
plt.show()
def prob_4_telescope():
# Repeat experiments for magic telescope and housing using weights (w = 1/dist**2)
train = Arff('datasets/magic_telescope_train.arff')
test = Arff('datasets/magic_telescope_test.arff')
train.normalize()
test.normalize()
krange = np.arange(1, 16, 2)
accs = []
for k in krange:
knn = KNN(k, weighting=True)
predictions = knn.knn(train.get_features(), train.get_labels(),
test.get_features())
acc = predictions == np.ravel(test.get_labels().data)
print("k:", k, "accuracy:", sum(acc) / len(acc))
accs.append(sum(acc) / len(acc))
plt.plot(krange, accs)
plt.title("K Size Versus Accuracy")
plt.xlabel("K")
plt.ylabel("Accuracy")
plt.show()
def prob_2():
# try first without normalizing
train = Arff('datasets/magic_telescope_train.arff')
test = Arff('datasets/magic_telescope_test.arff')
k = KNN(3)
predictions = k.knn(train.get_features(), train.get_labels(),
test.get_features())
acc = predictions == np.ravel(test.get_labels().data)
print("Before normalization:", sum(acc) / len(acc))
train.normalize()
test.normalize()
predictions = k.knn(train.get_features(), train.get_labels(),
test.get_features())
acc = predictions == np.ravel(test.get_labels().data)
print("After normalization:", sum(acc) / len(acc))
print("PART TWO:")
krange = np.arange(1, 16, 2)
accs = []
for k in krange:
knn = KNN(k)
predictions = knn.knn(train.get_features(), train.get_labels(),
test.get_features())
acc = predictions == np.ravel(test.get_labels().data)
print("k:", k, "accuracy:", sum(acc) / len(acc))
accs.append(sum(acc) / len(acc))
plt.plot(krange, accs)
plt.title("K Size Versus Accuracy")
plt.xlabel("K")
plt.ylabel("Accuracy")
plt.show()
def prob_4_housing():
# Repeat experiments for magic telescope and housing using weights (w = 1/dist**2)
train = Arff('datasets/housing_train.arff')
test = Arff('datasets/housing_test.arff')
train.normalize()
test.normalize()
krange = np.arange(1, 16, 2)
mses = []
for k in krange:
knn = KNN(k, weighting=True)
preds = knn.knn_regression(train.get_features(), train.get_labels(),
test.get_features())
mse = np.sum(
(preds - np.ravel(test.get_labels().data))**2, axis=0) / len(preds)
mses.append(mse)
plt.plot(krange, mses)
plt.title("K Size Versus MSE on Housing (Weighted)")
plt.xlabel("K")
plt.ylabel("Mean Squared Error")
plt.show()
def prob3_normalized():
arff = Arff('datasets/abalone.arff', label_count=0)
arff.normalize()
domain = np.arange(2, 8)
ssekmm = []
for k in domain:
km = KMeans(k)
ssek = km.train(arff)
ssekmm.append(ssek)
hac = HAC()
hac2 = HAC(simple=False)
ssehac = hac.train(arff, printk=domain)
ssehac2 = hac2.train(arff, printk=domain)
plt.plot(domain, ssekmm, label="K-Means SSE")
plt.plot(domain, ssehac[::-1], label="HAC (Single-Link) SSE")
plt.plot(domain, ssehac2[::-1], label="HAC (Complete-Link) SSE")
plt.title("Abalone SSE (Normalized) vs # of Clusters")
plt.xlabel("# of Clusters")
plt.ylabel('SSE')
plt.legend()
plt.show()
