def k_means_clustering_model(algo_type):
ds = st.sidebar.selectbox(
"Type of Dataset",
["Iris Flower Dataset", "Wine Quality Dataset", "Other"])
dataset = False
status = None
if ds == "Iris Flower Dataset":
dataset = True
status, X_train, Y_train, _, _, last = LOAD.load_internal_csv_dataset(
"Iris.csv", 0, "Clustering")
elif ds == "Wine Quality Dataset":
dataset = True
status, X_train, Y_train, _, _, last = LOAD.load_internal_csv_dataset(
"wine.csv", 0, "Clustering")
elif ds == "Other":
st.sidebar.markdown(
"""<ul><li>File Format: CSV</li><li>Must be preprocessed</li> <li>Must not contain any NAN values</li>
<li>Must not contain invalid values (Combination of numeric and non-numeric)</li>
<li>Assumption: First n-1 columns are numeric and last column (output) can be either numeric or non-numeric </li>""",
unsafe_allow_html=True)
dataset = st.sidebar.file_uploader("Upload Dataset")
status, X_train, Y_train, _, _, last = LOAD.load_external_csv_dataset(
dataset, 0, "Clustering")
if dataset and status == "VALID DATASET":
st.sidebar.success(status)
label = "Seeds"
X_train[label] = list(map(str, range(1, X_train.shape[0] + 1)))
st.sidebar.dataframe(X_train)
st.sidebar.text(f"Training data size: {X_train.shape}")
k = top.slider(f"Value of K ", 1, 100)
cluster_names = top.text_input(
"Enter valid {} cluster seed names:".format(k))
cluster_names = cluster_names.strip().split()
TB = top.button("Train")
if TB:
clustered_data = KMC.k_means_clustering(X_train, cluster_names,
label)
middle.dataframe(clustered_data)
clusters = clustered_data["cluster"].unique()
middle.write(f"{k} distinct clusters are : {clusters}")
plot(clustered_data, None, "cluster", algo_type)
st.balloons()
def multilayer_perceptron_model(algo_type):
top.markdown(
"""<p style ="font-style:italic;">Multilayer perceptron is a feedforward neural network that consists three types of layers
given below:</p> <ul><li>Input layer</li><li>Hidden layer</li><li>Output layer</li></ul><p style ="font-style:italic;">Input layer size (number of perceptrons in input layer)
must be equal to the number of features in dataset. Output layer size must be equal to the number of distinct classes in
the output vector.<br>Note: Hidden layer might have more than one layer in itself</p>""",
unsafe_allow_html=True)
top.info(
"Note: This Multilayer perceptron models performs only Binary classification"
)
ds = st.sidebar.selectbox("Type of Dataset", [
"XOR Gate", "Can vs Non cat image classification",
"Wine Quality Classification", "Heart Disease Chances Prediction",
"Other"
])
dataset = False
status = None
if ds == "XOR Gate" or ds == "Can vs Non cat image classification":
split = None
split_percent = True
else:
split = st.sidebar.text_input(
"Enter test split percentage between 1 and 100 ( e.g. 30 or 40)")
split_percent = None
if split:
try:
split_percent = int(split)
if split_percent < 1 or split_percent > 100:
st.stop()
except:
st.sidebar.error(
"SPLIT PERCENTAGE MUST BE AN INTEGER (BETWEEN 1 TO 100)")
st.stop()
if ds == "Can vs Non cat image classification" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test = LOAD.load_cat_noncat_dataset(
)
last = "category"
#flatten the data i.e change 12 ,12 , 3 array to (1,12*12*3)
m = X_train.shape[0]
n = X_test.shape[0]
X_train = X_train.reshape(m, -1)
X_test = X_test.reshape(n, -1)
Y_train = Y_train.reshape(1, m)
Y_test = Y_test.reshape(1, n)
classes = ["Cat", "NonCat"]
#normalizing the array
X_train = X_train / 255
X_test = X_test / 255
layers = None
elif ds == "XOR Gate" and split_percent:
dataset = True
n = top.slider("No. of Inputs", 2, 7)
status, X_train, Y_train, X_test, Y_test, last = LOAD.create_boolean_dataset(
n, "XOR")
elif ds == "Wine Quality Classification" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"wine.csv", split_percent, "Classification")
elif ds == "Heart Disease Chances Prediction" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"heart.csv", split_percent, "Classification")
elif ds == "Other" and split_percent:
st.sidebar.markdown(
"""<ul><li>File Format: CSV</li><li>Must be preprocessed</li> <li>Must not contain any NAN values</li>
<li>Must not contain invalid values (Combination of numeric and non-numeric)</li>
<li>Assumption: First n-1 columns are numeric and last column (output) can be either numeric or non-numeric </li>""",
unsafe_allow_html=True)
dataset = st.sidebar.file_uploader("Upload Dataset")
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_external_csv_dataset(
dataset, split_percent, "Classification")
if status == "VALID DATASET" and dataset:
st.sidebar.success(status)
train_data, test_data = X_train.copy(), X_test.copy()
if ds != "Can vs Non cat image classification":
train_data[last], test_data[last] = Y_train, Y_test
classes = list(test_data[last].unique())
st.sidebar.text(f"Classes are: {classes}")
st.sidebar.dataframe(train_data.head(10))
if len(classes) > 2:
top.error(
"More than two classes found in dataset. Can't classify")
st.stop()
st.sidebar.text(f"Training data size: {train_data.shape}")
st.sidebar.text(f"Testing data size: {test_data.shape}")
nl = top.text_input("Total number of layers in hidden layer")
no_l = None
try:
no_l = int(nl)
if no_l < 0:
top.error("Number of layers can not be less than 0")
st.stop()
except:
top.error("Number of layers must be positive integer")
st.stop()
if no_l is not None:
layers = [0] * (no_l + 2)
layers[0] = X_train.shape[1]
layers[-1] = 1
top.text(f"Input layer size : {layers[0]}")
for i in range(1, no_l + 1):
layers[i] = top.text_input(f"Hidden layer{i} size : ")
try:
layers[i] = int(layers[i])
if layers[i] < 0:
top.error("Size of layer can not be less than 0")
break
except:
top.error("Size of layers must be positive integer")
break
top.text(f"Output layer size: {layers[-1]}")
input_activation = top.selectbox(
"Activation function for first n-1 layers",
["relu", "sigmoid", "tanh"])
output_activation = top.selectbox(
"Activation function for the ouput layer", ["sigmoid"])
learning_rate = top.number_input("Learning Rate")
num_iterations = top.slider("Iterations", 1, 10001)
top.markdown(
"""<ul style="text-align:justify;"><li>Default activation functions: Relu and Sigmoid respectively (recommended)</li>
<li>Learning rate is decreasing as number of iterations increases. (for e.g. if steady learning rate = 0.1
; total iterations=100 then dynamic learning rate = 0.1/100</li> </ul>""",
unsafe_allow_html=True)
TB = top.button("Train")
if TB:
if ds != "Can vs Non cat image classification":
train_data[last], test_data[last] = pd.Categorical(
train_data[last]), pd.Categorical(test_data[last])
cat_map = dict(enumerate(test_data[last].cat.categories))
train_data[last], test_data[last] = train_data[
last].cat.codes, test_data[last].cat.codes
middle.text(cat_map)
Y_train, Y_test = np.array(
train_data[last].copy()), np.array(
test_data[last].copy())
Y_train, Y_test = Y_train.reshape(
1, train_data.shape[0]), Y_test.reshape(
1, test_data.shape[0])
X_train, X_test = DNN.normalize_data(X_train, X_test)
parameters, costs = DNN.deep_neural_network(
layers, learning_rate, num_iterations, X_train.T, Y_train,
input_activation, output_activation)
Z_train, accuracy, A_train = DNN.predict(
X_train.T, Y_train, parameters, input_activation,
output_activation)
middle.text("Training Accuracy : {:.2%}".format(accuracy))
Z_test, accuracy, A_test = DNN.predict(X_test.T, Y_test,
parameters,
input_activation,
output_activation)
middle.text("Testing Accuracy : {:.2%}".format(accuracy))
if ds != "Can vs Non cat image classification":
train_data["Prediction for " + last] = np.squeeze(Z_train)
test_data["Prediction for " + last] = np.squeeze(Z_test)
middle.dataframe(test_data)
plot_confusion_matrix(test_data[last],
test_data["Prediction for " + last])
else:
actual = list(np.squeeze(Y_test))
predicted = list(np.squeeze(Z_test))
for i in range(len(actual)):
actual[i] = "NonCat" if actual[i] == 0 else "Cat"
for i in range(len(predicted)):
predicted[i] = "NonCat" if predicted[i] == 0 else "Cat"
fig, ax = plt.subplots(1, 3)
for i in range(3):
img = test_data[i]
ax[i].imshow(img.reshape(64, 64, 3))
middle.text(f"Actual Category : {actual[:3]}")
middle.text(f"Predicted Category: {predicted[:3]}")
middle.pyplot(fig)
data = {'Actual': actual, 'Prediction': predicted}
df = pd.DataFrame(data)
plot_confusion_matrix(df['Actual'], df['Prediction'])
plot_cost(costs)
st.balloons()
elif status and dataset:
st.sidebar.error(status)
st.stop()
def logistic_regression_model(algo_type):
top.info(
"Note: This Logistic Regression models performs only Binary classification"
)
ds = st.sidebar.selectbox(
"Choose Dataset (Choose Other to Upload External File)", [
"Logic Gates", "Wine Quality Classification",
"Heart Disease Chances Prediction", "Other"
])
dataset = False
status = None
split_percent = None
if ds == "Logic Gates":
split = None
split_percent = True
else:
split = st.sidebar.text_input(
"Enter test split percentage between 1 and 100 ( e.g. 30 or 40)")
if split:
try:
split_percent = int(split)
if split_percent < 1 or split_percent > 100:
st.stop()
except:
st.sidebar.error(
"SPLIT PERCENTAGE MUST BE AN INTEGER (BETWEEN 1 TO 100)")
st.stop()
if ds == "Logic Gates" and split_percent:
dataset = True
gate = top.selectbox("Choose Gate", ["AND", "OR", "NAND", "NOR"])
n = top.selectbox("Number of Inputs", range(2, 7))
status, X_train, Y_train, X_test, Y_test, last = LOAD.create_boolean_dataset(
n, gate)
elif ds == "Wine Quality Classification" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"wine.csv", split_percent, "Classification")
elif ds == "Heart Disease Chances Prediction" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"heart.csv", split_percent, "Classification")
elif ds == "Other" and split_percent:
st.sidebar.markdown(
"""<ul><li>File Format: CSV</li><li>Must be preprocessed</li> <li>Must not contain any NAN values</li>
<li>Must not contain invalid values (Combination of numeric and non-numeric)</li>
<li>Assumption: First n-1 columns are numeric and last column (output) can be either numeric or non-numeric </li>""",
unsafe_allow_html=True)
dataset = st.sidebar.file_uploader("Upload Dataset")
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_external_csv_dataset(
dataset, split_percent, "Classification")
if dataset and status == "VALID DATASET":
st.sidebar.success(status)
train_data, test_data = X_train.copy(), X_test.copy()
train_data[last], test_data[last] = Y_train, Y_test
classes = list(test_data[last].unique())
st.sidebar.text(f"Classes are: {classes}")
st.sidebar.dataframe(test_data.head(10))
if len(classes) > 2:
top.error("More than two classes found in dataset. Can't classify")
st.stop()
st.sidebar.text(f"Training data size: {train_data.shape}")
st.sidebar.text(f"Testing data size: {test_data.shape}")
learning_rate = top.number_input("Learning Rate")
num_iterations = top.slider("Iterations", 1, 5000)
TB = top.button("Train")
top.markdown(
"""<ul style="text-align:justify;"><li>Learning is decreasing as number of iterations increases (for e.g. if steady learning rate = 0.1
; total iterations=100 then dynamic learning rate = 0.1/100</li> </ul>""",
unsafe_allow_html=True)
if TB and len(classes) <= 2:
train_data[last], test_data[last] = pd.Categorical(
train_data[last]), pd.Categorical(test_data[last])
cat_map = dict(enumerate(test_data[last].cat.categories))
train_data[last], test_data[last] = train_data[
last].cat.codes, test_data[last].cat.codes
middle.text(cat_map)
Y_train, Y_test = np.array(train_data[last].copy()), np.array(
test_data[last].copy())
Y_train, Y_test = Y_train.reshape(
1, train_data.shape[0]), Y_test.reshape(1, test_data.shape[0])
#if ouptut vector contains string
X_train = Logistic.normalize_data(X_train)
X_test = Logistic.normalize_data(X_test)
w, b, costs = Logistic.logistic_regression(X_train.T, Y_train,
learning_rate,
num_iterations)
Z_train, accuracy, A_train = Logistic.predict(
X_train.T, Y_train, w, b)
middle.text("Training Accuracy : {:.2%}".format(accuracy))
Z_test, accuracy, A_test = Logistic.predict(X_test.T, Y_test, w, b)
middle.text("Testing Accuracy : {:.2%}".format(accuracy))
train_data["Prediction for " + last] = np.squeeze(Z_train)
test_data["Prediction for " + last] = np.squeeze(Z_test)
middle.dataframe(test_data)
plot_confusion_matrix(test_data[last],
test_data["Prediction for " + last])
plot_decision_boundary(test_data, w, b, last)
plot_cost(costs)
st.balloons()
elif status and dataset:
st.sidebar.error(status)
st.stop()
def linear_regression_model(algo_type):
options = [
"Father Son Height Prediction", "Graduate Admission Chance Prediction",
"Other"
]
ds = st.sidebar.selectbox(
"Choose Dataset (Choose Other to Upload External File)", options)
split = st.sidebar.text_input(
"Enter test split percentage between 1 and 100 ( e.g. 30 or 40)")
split_percent = None
dataset = False
status = None
if split:
try:
split_percent = int(split)
if split_percent < 1 or split_percent > 100:
st.stop()
except:
st.sidebar.error(
"SPLIT PERCENTAGE MUST BE AN INTEGER (BETWEEN 1 TO 100)")
st.stop()
if ds == "Father Son Height Prediction" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"father_son_height.csv", split_percent, "Regression")
elif ds == "Graduate Admission Chance Prediction" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"Admission_predict.csv", split_percent, "Regression")
elif ds == "Other" and split_percent:
html = """<ul><li>File Format: CSV</li><li>Must be preprocessed</li> <li>Must not contain any NAN values</li>
<li>Must not contain invalid values (Combination of numeric and non-numeric)</li>
<li>Assumption: First n-1 columns are numeric and last column (output) can be either numeric or non-numeric </li></ul>"""
st.sidebar.markdown(html, unsafe_allow_html=True)
dataset = st.sidebar.file_uploader("Upload Dataset")
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_external_csv_dataset(
dataset, split_percent, "Regression")
if status == "VALID DATASET" and dataset:
st.sidebar.success(status)
train_data, test_data = X_train.copy(), X_test.copy()
train_data[last], test_data[last] = Y_train, Y_test
st.sidebar.dataframe(train_data.head(10))
st.sidebar.text(f"Training data size: {train_data.shape}")
st.sidebar.text(f"Testing data size: {test_data.shape}")
learning_rate = top.number_input("Learning Rate")
num_iterations = top.slider("Iterations", 1, 5000)
TB = top.button("Train")
top.markdown(
"""<ul style="text-align:justify;"><li>Learning is decreasing as number of
iterations increases (for e.g. if steady learning rate = 0.1 & total iterations=100
then dynamic learning rate = 0.1/100</li> </ul>""",
unsafe_allow_html=True)
if TB:
Y_train, Y_test = np.array(Y_train.copy()), np.array(Y_test.copy())
Y_train = Y_train.reshape(1, train_data.shape[0])
Y_test = Y_test.reshape(1, test_data.shape[0])
X_train = Linear.normalize_data(X_train)
X_test = Linear.normalize_data(X_test)
w, b, costs = Linear.linear_regression(X_train.T, Y_train,
learning_rate,
num_iterations)
Z_train = np.dot(w, X_train.T) + b
MSE = np.sum((Z_train - Y_train)**2) / Y_train.shape[1]
middle.text(f"Training Mean squared error is: {MSE}")
Z_test = np.dot(w, X_test.T) + b
MSE = np.sum((Z_test - Y_test)**2) / Y_test.shape[1]
middle.text(f"Testing Mean squared error is: {MSE}")
train_data["Prediction for " + last] = np.round(
np.squeeze(Z_train), 2)
test_data["Prediction for " + last] = np.round(
np.squeeze(Z_test), 2)
middle.dataframe(train_data)
plot(train_data, test_data, last, algo_type)
plot_cost(costs)
st.balloons()
elif status is not None and dataset:
st.sidebar.error(status)
st.stop()
def naive_bayes_model(algo_type):
ds = st.sidebar.selectbox(
"Choose Dataset (Choose Other to Upload External File)", [
"Iris Flower Classification", "Wine Quality Classification",
"Heart Disease Chances Prediction", "Other"
])
split = st.sidebar.text_input(
"Enter test split percentage between 1 and 100 ( e.g. 30 or 40)")
split_percent = None
dataset = False
status = None
if split:
try:
split_percent = int(split)
if split_percent < 1 or split_percent > 100:
st.stop()
except:
st.sidebar.error(
"SPLIT PERCENTAGE MUST BE AN INTEGER (BETWEEN 1 TO 100)")
st.stop()
if ds == "Iris Flower Classification" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"Iris.csv", split_percent, "Classification")
elif ds == "Wine Quality Classification" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"wine.csv", split_percent, "Classification")
elif ds == "Heart Disease Chances Prediction" and split_percent:
dataset = True
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_internal_csv_dataset(
"heart.csv", split_percent, "Classification")
elif ds == "Other" and split_percent:
st.sidebar.markdown(
"""<ul><li>File Format: CSV</li><li>Must be preprocessed</li> <li>Must not contain any NAN values</li>
<li>Must not contain invalid values (Combination of numeric and non-numeric)</li>
<li>Assumption: First n-1 columns are numeric and last column (output) can be either numeric or non-numeric </li>""",
unsafe_allow_html=True)
dataset = st.sidebar.file_uploader("Upload Dataset")
status, X_train, Y_train, X_test, Y_test, last = LOAD.load_external_csv_dataset(
dataset, split_percent, "Classification")
if status == "VALID DATASET" and dataset:
st.sidebar.success(status)
train_data, test_data = X_train.copy(), X_test.copy()
train_data[last], test_data[last] = Y_train, Y_test
classes = list(train_data[last].unique())
st.sidebar.dataframe(train_data.head(10))
st.sidebar.text(f"Training data size: {train_data.shape}")
st.sidebar.text(f"Testing data size: {test_data.shape}")
TB = top.button("Train")
if TB:
marginal_prob, mean_std = NB.gaussian_naive_bayes_classiefier(
train_data, classes, last)
Z_train, accuracy = NB.predict(X_train, Y_train, mean_std,
marginal_prob, classes)
middle.text("Training Accuracy : {:.2%}".format(accuracy))
Z_test, accuracy = NB.predict(X_test, Y_test, mean_std,
marginal_prob, classes)
middle.text("Testing Accuracy : {:.2%}".format(accuracy))
test_data["Prediction for " + last] = np.squeeze(Z_test)
train_data["Prediction for " + last] = np.squeeze(Z_train)
middle.dataframe(train_data.head(10))
plot_confusion_matrix(test_data[last],
test_data["Prediction for " + last])
st.balloons()
elif status and dataset:
st.sidebar.error(status)
st.stop()