Scribd
Upload
5 views12 pages

Project Code

The document outlines a series of tasks involving the Iris dataset, focusing on visualizing decision boundaries using logistic regression with and without polynomial features. It includes data preprocessing steps like scaling and splitting, as well as plotting techniques for decision boundaries, class distributions, and feature distributions. Additionally, it demonstrates the use of sigmoid functions and correlation heatmaps to analyze the dataset.

Uploaded by

Haroon Zafar
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views12 pages

Project Code

The document outlines a series of tasks involving the Iris dataset, focusing on visualizing decision boundaries using logistic regression with and without polynomial features. It includes data preprocessing steps like scaling and splitting, as well as plotting techniques for decision boundaries, class distributions, and feature distributions. Additionally, it demonstrates the use of sigmoid functions and correlation heatmaps to analyze the dataset.

Uploaded by

Haroon Zafar
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Project Code

# Import libraries

import numpy as np

import [Link] as plt

from [Link] import load_iris

from sklearn.model_selection import train_test_split

from [Link] import StandardScaler

# Task 1

# Use only first two features for visualization

# Sepal length & Sepal width

iris = load_iris()

X = [Link][:, :2]

y = [Link]

# Data Splitting

X_train, X_test, y_train, y_test = train_test_split(

X, y, test_size=0.20, random_state=0

# features Scaling

sc = StandardScaler()

X_train = sc.fit_transform(X_train)

X_test = [Link](X_test)

# perceptron trick

from [Link] import PolynomialFeatures

poly = PolynomialFeatures(degree=2)

X_train_poly = poly.fit_transform(X_train)

X_test_poly = [Link](X_test)

# Training the Model

from sklearn.linear_model import LogisticRegression

classifier = LogisticRegression(multi_class='ovr', solver='lbfgs', random_state=0)

[Link](X_train_poly, y_train)
# Mesh Grid

x_min, x_max = X_train[:, 0].min() - 1, X_train[:, 0].max() + 1

y_min, y_max = X_train[:, 1].min() - 1, X_train[:, 1].max() + 1

x, y = [Link](

[Link](x_min, x_max, 0.01),

[Link](y_min, y_max, 0.01)

grid = np.c_[[Link](), [Link]()]

grid_poly = [Link](grid)

# Decision boudary plotting

Z = [Link](grid_poly)

Z = [Link]([Link])

[Link](figsize=(8,6))

[Link](x, y, Z, alpha=0.5)

colors = ['red', 'green', 'blue']

labels = iris.target_names

for i, color, label in zip(range(3), colors, labels):

[Link](

X_train[y_train == i, 0],

X_train[y_train == i, 1],

c=color, label=label, edgecolor='k'

[Link]('Sepal Length (scaled)')

[Link]('Sepal Width (scaled)')

[Link]('Decision Boundary for Iris Classes with percepron trick')

[Link]()

[Link]()

#Task 2

# Use only first two features for visualization


# Sepal length & Sepal width

iris = load_iris()

X = [Link][:, :2]

y = [Link]

# Data Splitting

X_train, X_test, y_train, y_test = train_test_split(

X, y, test_size=0.20, random_state=0

# features Scaling

sc = StandardScaler()

X_train = sc.fit_transform(X_train)

X_test = [Link](X_test)

# Training the Model without perceptron trick

from sklearn.linear_model import LogisticRegression

classifier = LogisticRegression(multi_class='ovr', solver='lbfgs', random_state=0)

[Link](X_train, y_train)

# Mesh Grid

x_min, x_max = X_train[:, 0].min() - 1, X_train[:, 0].max() + 1

y_min, y_max = X_train[:, 1].min() - 1, X_train[:, 1].max() + 1

x, y = [Link](

[Link](x_min, x_max, 0.01),

[Link](y_min, y_max, 0.01)

grid = np.c_[[Link](), [Link]()]

# Decision boudary plotting

Z2 = [Link](grid)

Z2 = [Link]([Link])
[Link](figsize=(8,6))

[Link](x, y, Z2, alpha=0.5)

colors = ['red', 'green', 'blue']

labels = iris.target_names

for i, color, label in zip(range(3), colors, labels):

[Link](

X_train[y_train == i, 0],

X_train[y_train == i, 1],

c=color, label=label, edgecolor='k'

[Link]('Sepal Length (scaled)')

[Link]('Sepal Width (scaled)')

[Link]('Decision Boundary with only logistic Regression')

[Link]()

[Link]()

# Task 3

# sigmoid function

# Use only first two features for visualization

# Sepal length & Sepal width

iris = load_iris()

X = [Link][:, :2]

y = [Link]

# Data Splitting

X_train, X_test, y_train, y_test = train_test_split(


X, y, test_size=0.20, random_state=0

# features Scaling

sc = StandardScaler()

X_train = sc.fit_transform(X_train)

X_test = [Link](X_test)

# Training the Model

from sklearn.linear_model import LogisticRegression

classifier = LogisticRegression(multi_class='ovr', random_state=0)

[Link](X_train, y_train)

def sigmoid(z1):

return 1 / (1 + [Link](-z1))

# Create mesh grid

x_min, x_max = X[:, 0].min() - 1, X[:, 0].max() + 1

y_min, y_max = X[:, 1].min() - 1, X[:, 1].max() + 1

xx, yy = [Link](

[Link](x_min, x_max, 300),

[Link](y_min, y_max, 300)


)

# Compute decision scores

grid = np.c_[[Link](), [Link]()]

grid = [Link](grid)

z1 = classifier.decision_function(grid)

# Convert scores to probabilities using sigmoid

probabilities = sigmoid(z1)

# Pick class with highest probability

z1 = [Link](probabilities, axis=1)

z1 = [Link]([Link])

# Plot decision boundary

[Link](figsize=(8, 6))

[Link](xx, yy, z1, alpha=0.5)

[Link](X[:, 0], X[:, 1], c=y, edgecolor='k')

[Link]("Sepal Length")

[Link]("Sepal Width")

[Link]("Decision Boundary of iris using Sigmoid)")

[Link]()

[Link]()

# Task 4

# Import libraries

import numpy as np

import [Link] as plt

from [Link] import load_iris


from sklearn.model_selection import train_test_split

from [Link] import StandardScaler, PolynomialFeatures

from sklearn.linear_model import LogisticRegression

# Load dataset (2 features only)

iris = load_iris()

X = [Link][:, :2]

y = [Link]

# Train-test split

X_train, X_test, y_train, y_test = train_test_split(

X, y, test_size=0.20, random_state=0

# Feature scaling

sc = StandardScaler()

X_train_scaled = sc.fit_transform(X_train)

X_test_scaled = [Link](X_test)

# Perceptron trick (Polynomial Features)

poly = PolynomialFeatures(degree=2)

X_train_poly = poly.fit_transform(X_train_scaled)

# Train Logistic Regression (OvR)

classifier = LogisticRegression(multi_class='ovr', random_state=0)

[Link](X_train_poly, y_train)

# Sigmoid function

def sigmoid(z):

return 1 / (1 + [Link](-z))
# Mesh grid (in scaled feature space)

x_min, x_max = X_train_scaled[:, 0].min() - 1, X_train_scaled[:, 0].max() + 1

y_min, y_max = X_train_scaled[:, 1].min() - 1, X_train_scaled[:, 1].max() + 1

xx, yy = [Link](

[Link](x_min, x_max, 300),

[Link](y_min, y_max, 300)

# Prepare grid

grid = np.c_[[Link](), [Link]()]

grid_poly = [Link](grid)

# Decision function → Sigmoid

scores = classifier.decision_function(grid_poly)

probabilities = sigmoid(scores)

# Final class prediction

Z = [Link](probabilities, axis=1)

Z = [Link]([Link])

# Plot decision boundary

[Link](figsize=(8, 6))

[Link](xx, yy, Z, alpha=0.5)

colors = ['red', 'green', 'blue']

labels = iris.target_names

for i, color, label in zip(range(3), colors, labels):

[Link](

X_train_scaled[y_train == i, 0],
X_train_scaled[y_train == i, 1],

c=color, label=label, edgecolor='k'

[Link]("Sepal Length (scaled)")

[Link]("Sepal Width (scaled)")

[Link]("Decision Boundary using Perceptron Trick + Sigmoid ")

[Link]()

[Link]()

# Pie Chart (Class Distribution)

import [Link] as plt

from [Link] import load_iris

import numpy as np

iris = load_iris()

y = [Link]

labels = iris.target_names

sizes = [Link](y)

[Link](figsize=(6,6))

[Link](sizes, labels=labels, autopct='%1.1f%%', startangle=90)

[Link]("Pie Chart of Iris Class Distribution")

[Link]()

# Histogram (Feature Distribution)

import [Link] as plt

from [Link] import load_iris

iris = load_iris()
X = [Link]

feature_names = iris.feature_names

[Link](figsize=(8,5))

[Link](X[:, 0], bins=20, alpha=0.7)

[Link](feature_names[0])

[Link]("Frequency")

[Link]("Histogram of Sepal Length")

[Link]()

# Gaussian Distribution (Normal Curve)

import numpy as np

import [Link] as plt

from [Link] import load_iris

from [Link] import norm

iris = load_iris()

data = [Link][:, 0] # Sepal length

mean, std = [Link](), [Link]()

x = [Link]([Link](), [Link](), 100)

pdf = [Link](x, mean, std)

[Link](figsize=(8,5))

[Link](data, bins=20, density=True, alpha=0.6)

[Link](x, pdf)

[Link]("Sepal Length")

[Link]("Probability Density")

[Link]("Gaussian Distribution of Sepal Length")

[Link]()
# Correlation Heatmap

import seaborn as sns

import [Link] as plt

import pandas as pd

from [Link] import load_iris

iris = load_iris()

df = [Link]([Link], columns=iris.feature_names)

corr = [Link]()

[Link](figsize=(8,6))

[Link](corr, annot=True, cmap='coolwarm')

[Link]("Correlation Heatmap of Iris Features")

[Link]()

You might also like