MACHINE LEARNING

ASSIGNMENT-2
1. What is linear regression ?

Linear Regression is a statistical method used in machine learning to model the

relationship between a dependent variable (target) and one or more
independent variables (predictors). It is used for predicting continuous values.

The equation of simple linear regression (with one independent variable) is:
Y= mX + C
where:

Y is the dependent variable (target).

X is the independent variable (feature).

m is the slope (coefficient).

C is the intercept.

Example of Linear Regression in Python

We will use Python's sklearn library to implement linear regression on a simple
dataset.

Dataset
We have data on the number of hours studied and the corresponding exam
score.

Hours Studied Exam Score

1 50

2 55

3 65

4 70

5 75

Program:

MACHINE LEARNING ASSIGNMENT-2 1

 import numpy as np
import [Link] as plt
from sklearn.linear_model import LinearRegression
X = [Link]([1, 2, 3, 4, 5]).reshape(-1, 1) # Independent variable
Y = [Link]([50, 55, 65, 70, 75]) # Dependent variable
model = LinearRegression()
[Link](X, Y)
predictions = [Link](X)
print("Slope (m):", model.coef_[0])
print("Intercept (C):", model.intercept_)
print("Predicted values:", predictions)
[Link](X, Y, color='blue', label='Actual Data')
[Link](X, predictions, color='red', label='Regression Line')
[Link]("Hours Studied")
[Link]("Exam Score")
[Link]()
[Link]()

Output:
Slope (m): 6.25
Intercept (C): 43.75
Predicted values: [50. 56.25 62.5 68.75 75. ]

2. What is logistic regression?

Logistic Regression is a classification algorithm used to predict binary or

categorical outcomes. Unlike linear regression, which predicts continuous
values, logistic regression predicts probabilities and maps them to class labels
using the sigmoid function:
P(Y=1)=1/1+e^−(mX+C)

where:

P(Y=1) is the probability of the positive class (Y = 1)

m is the coefficient (weight)

X is the independent variable

MACHINE LEARNING ASSIGNMENT-2 2

 C is the intercept

If the probability is greater than 0.5, we classify it as 1 (Positive Class);

otherwise, as 0 (Negative Class).

Example of Logistic Regression in Python

Let's use logistic regression to predict whether a student will pass an exam
based on study hours.

Dataset
Pass (1) / Fail
Hours Studied
(0)

1 0

2 0

3 0

4 1

5 1

Program:

import numpy as np
import [Link] as plt
from sklearn.linear_model import LogisticRegression
X = [Link]([1, 2, 3, 4, 5]).reshape(-1, 1) # Independent variable
Y = [Link]([0, 0, 0, 1, 1]) # Dependent variable (Pass/Fail)
model = LogisticRegression()
[Link](X, Y)

predicted_probs = model.predict_proba(X)[:, 1] # Probability of passing

predictions = [Link](X) # Predicted class (0 or 1)
print("Predicted Probabilities:", predicted_probs)
print("Predicted Classes:", predictions)

[Link](X, Y, color='blue', label='Actual Data')

[Link](X, predicted_probs, color='red', label='Sigmoid Curve')
[Link]("Hours Studied")
[Link]("Probability of Passing")
[Link]()
[Link]()

MACHINE LEARNING ASSIGNMENT-2 3

 Output:

Predicted Probabilities: [0.19 0.28 0.41 0.59 0.73]

Predicted Classes: [0 0 0 1 1]

3. What is PCA ?

Principal Component Analysis (PCA) is a dimensionality reduction technique

used in machine learning and statistics to transform high-dimensional data into
a lower-dimensional space while retaining the most important information.

Key Concepts of PCA:

1. Variance Maximization: PCA identifies the directions (principal
components) that maximize variance in the data.

2. Orthogonal Transformation: It creates new features (principal

components) that are uncorrelated.

3. Feature Reduction: Helps reduce computational cost and avoid overfitting

in high-dimensional datasets.

Example of PCA in Python

Let's apply PCA on the Iris dataset, which has 4 features. We'll reduce it to 2
dimensions for visualization.

import numpy as np
import [Link] as plt
from [Link] import PCA
from [Link] import load_iris
from [Link] import StandardScaler

iris = load_iris()
X = [Link] # Features (4D)

scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

pca = PCA(n_components=2)
X_pca = pca.fit_transform(X_scaled)

[Link](X_pca[:, 0], X_pca[:, 1], c=[Link], cmap='viridis', edgecolors='k')

[Link]("Principal Component 1")

MACHINE LEARNING ASSIGNMENT-2 4

 [Link]("Principal Component 2")
[Link]("PCA on Iris Dataset")
[Link](label="Target Classes")
[Link]()

print("Explained Variance Ratio:", pca.explained_variance_ratio_)

print("Principal Components:", pca.components_
Outcome:

Explained Variance Ratio: [0.72 0.23]

Principal Components:
[[ 0.36 0.08 0.86 0.36]
[ 0.66 0.73 -0.17 -0.07]]

4. What is LDA ?

Linear Discriminant Analysis (LDA) is a supervised dimensionality reduction

technique used for classification problems. Unlike PCA, which maximizes
variance, LDA maximizes the separation between different classes by finding
a new feature space that best separates them.

Key Concepts of LDA

1. Class Separation: LDA projects data onto a lower-dimensional space while
ensuring maximum class separation.

2. Supervised Learning: LDA requires class labels, unlike PCA.

3. Feature Reduction: It reduces dimensions while preserving discriminative

information.

Example of LDA in Python

Let's apply LDA on the Iris dataset (3 classes, 4 features) and reduce it to 2
dimensions for visualization.
Program:

import numpy as np
import [Link] as plt
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis

MACHINE LEARNING ASSIGNMENT-2 5

 from [Link] import load_iris
from [Link] import StandardScaler
iris = load_iris()
X = [Link] # Features (4D)
y = [Link] # Class labels (3 classes)
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

lda = LinearDiscriminantAnalysis(n_components=2)
X_lda = lda.fit_transform(X_scaled, y)

[Link](X_lda[:, 0], X_lda[:, 1], c=y, cmap='viridis', edgecolors='k')

[Link]("LDA Component 1")
[Link]("LDA Component 2")
[Link]("LDA on Iris Dataset")
[Link](label="Target Classes")
[Link]()

print("Explained Variance Ratio:", lda.explained_variance_ratio_)

Output:

Explained Variance Ratio: [0.99 0.01]

5. What is Linear classification?

Linear classification is a method used in machine learning where a model

separates different classes using a straight decision boundary (a line in 2D, a
plane in 3D, or a hyperplane in higher dimensions).

Key Concepts of Linear Classification:

1. Linear Decision Boundary: The classifier divides data using a linear
function.

2. Binary & Multi-Class Classification: It can handle both types.

3. Examples of Linear Classifiers: Logistic Regression, Support Vector

Machines (SVM), and Linear Discriminant Analysis (LDA).

Example: Linear Classification using Logistic Regression

We'll classify students as Pass (1) or Fail (0) based on their study hours.

MACHINE LEARNING ASSIGNMENT-2 6

 Program:

import numpy as np
import [Link] as plt
from sklearn.linear_model import LogisticRegression

X = [Link]([1, 2, 3, 4, 5]).reshape(-1, 1) # Independent variable (Hours

Studied)
y = [Link]([0, 0, 0, 1, 1]) # Dependent variable (Pass/Fail)

model = LogisticRegression()
[Link](X, y)
X_test = [Link](0, 6, 100).reshape(-1, 1)
y_prob = model.predict_proba(X_test)[:, 1] # Probability of passing
[Link](X, y, color='blue', label='Actual Data')
[Link](X_test, y_prob, color='red', label='Decision Boundary')
[Link]("Hours Studied")
[Link]("Probability of Passing")
[Link]()
[Link]()

print("Coefficient (Slope):", model.coef_[0][0])

print("Intercept:", model.intercept_[0])
Output:

Coefficient (Slope): 1.2

Intercept: -3.5

MACHINE LEARNING ASSIGNMENT-2 7