def run_dr_clustering():
deposit_data = load_cleanse_data()
income_data = loadData()
pcakMeans(deposit_data, 7, 'deposit', 'manhattan')
# running PCA/kmeans for income Data
pcakMeans(income_data, 4, 'income', 'euclidean')
# running PCA/em for Deposit Data
pcaem(deposit_data, 7, 'deposit', 'manhattan')
# running PCA/em for Deposit Data
pcaem(income_data, 4, 'income', 'euclidean')
# ICA for deposit clustering
icakMeans(deposit_data, 35, 'deposit', 'manhattan')
icaem(deposit_data, 35, 'deposit', 'manhattan')
# ICA for income clustering
icakMeans(income_data, 12, 'income', 'euclidean')
icaem(income_data, 12, 'income', 'euclidean')
rpkMeans(deposit_data, 30, 'deposit', 'manhattan')
rpem(deposit_data, 30, 'deposit', 'manhattan')
# ICA for income clustering
rpkMeans(income_data, 8, 'income', 'euclidean')
rpem(income_data, 8, 'income', 'euclidean')
uvfskMeans(deposit_data, 30, 'deposit', 'manhattan')
uvfsem(deposit_data, 30, 'deposit', 'manhattan')
uvfskMeans(income_data, 10, 'income', 'euclidean')
uvfsem(income_data, 10, 'income', 'euclidean')
def performKmeansNN():
# load deposit dataset
data = load_cleanse_data()
# Select a range of k to check
target_clusters = [7, 10, 15, 20, 25, 30, 35, 40, 41]
mlp = MLPClassifier(hidden_layer_sizes=(15, 2),
random_state=70,
activation='relu',
max_iter=500)
scoring = ['accuracy']
scores = cross_validate(mlp,
data['features'],
data['labels'],
scoring=scoring,
cv=10)
print(scores)
NN_fit_time = np.mean(scores['fit_time'])
NN_accuracy = np.mean(scores['test_accuracy'])
kmeans_nn_accuracy = []
kmeans_nn_time = []
for cluster in target_clusters:
kmeans = KMeans(n_clusters=cluster, random_state=42)
clusters = kmeans.fit_predict(data['features'])
scores = cross_validate(mlp,
clusters.reshape(-1, 1),
data['labels'],
scoring=scoring,
cv=10)
kmeans_nn_accuracy.append(np.mean(scores['test_accuracy']))
kmeans_nn_time.append(np.mean(scores['fit_time']))
print(kmeans_nn_accuracy)
print(kmeans_nn_time)
plt.style.use("seaborn")
plt.figure(figsize=(8, 8))
plt.plot(target_clusters, kmeans_nn_accuracy)
plt.xticks(target_clusters)
plt.axhline(y=NN_accuracy, color='r', linestyle='-')
plt.xlabel("# Clusters")
plt.ylabel('NN Accuracy')
plt.grid(True)
plt.savefig('plots/kmeans_nn/deposit/kmeans_nn_accuracy.png')
plt.clf()
plt.style.use("seaborn")
plt.plot(target_clusters, kmeans_nn_time)
plt.xticks(target_clusters)
plt.axhline(y=NN_fit_time, color='r', linestyle='-')
plt.xlabel("Principal Components")
plt.ylabel('NN Fit Time')
plt.grid(True)
plt.savefig('plots/kmeans_nn/deposit/kmeans_nn_fit_time.png')
plt.clf()
def performDepositPCA():
data = load_cleanse_data()
pca.perform_pca(data['features'], 'deposit')
pca.validate_pca_nn(data, [7, 10, 15, 20, 25, 30, 35, 40, 41], 'deposit')
def performDeposituvfs():
data = load_cleanse_data()
uvfs.validate_uvfs_nn(data, [7, 10, 15, 20, 25, 30, 35, 40, 41],'deposit')
def performDepositRandomProjection():
data = load_cleanse_data()
randomprojection.apply_rp(data, [7, 10, 15, 20, 25, 30, 35, 40, 41],'deposit',2,3)
randomprojection.validate_rp_nn(data, [7, 10, 15, 20, 25, 30, 35, 40, 41],'deposit')
def deposit_clustering():
data = load_cleanse_data()
estimate_k(data, 'deposit', 'plots/kmeans/', 'manhattan')
validate_k(data, 'plots/kmeans/', 'deposit')
def deposit_em():
data = load_cleanse_data()
estimate_em_k(data, 'deposit', 'plots/em/', 'manhattan')
validate_em_k(data, 'plots/em/', 'deposit')