Scribd
Upload
7 views

Assignment-1 ML

Uploaded by

pnishith215
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
7 views

Assignment-1 ML

Uploaded by

pnishith215
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Machine Learning (CS-401T)

Assignment-1
Submission Deadline: 17/10/2024

1. Define Logistic Regression. Explain how it differs from linear regression.


Consider the following dataset representing whether a student will pass or fail an exam based on their study hours:
Study Hours Pass (1) / Fail (0)
1 0
2 0
3 0
4 1
5 1
6 1
7 1
8 1

Using the dataset, perform the following:


i) Fit a logistic regression model using a suitable library (e.g., Scikit-learn) and obtain the coefficients.
ii) Calculate the predicted probabilities of passing the exam for each study hour.
iii) Determine the classification outcome (Pass/Fail) using a threshold of 0.5.

2. Discuss the structure of an ANN, explaining the roles of the input layer, hidden layers, and output layer, Activation
functions in processing information.
Consider a 2-layer feed forward neural network with two input neurons, one hidden layer consisting of two
neurons, and one output neuron. The network uses the sigmoid activation function at both layers.
The inputs, weights, and biases are as follows:
Input vector: X=[0.6,0.9]
Weights for the hidden layer: Wh1=[0.3,0.5],Wh2=[0.4,0.7]
Weights for the output layer: Wo=[0.6,0.8]
Biases for the hidden neurons: bh1=0, bh2=0.
Bias for the output neuron: bo=0.4

a) Perform the feed-forward pass, showing all calculations for the activations of the hidden layer neurons and
the output neuron. Use the Sigmoid activation function.
b) b) Briefly explain how the back-propagation algorithm works to adjust the weights and biases of the network.
3. What is the Gradient Descent algorithm? Describe its purpose in machine learning. Consider a simple linear
regression model where the goal is to fit a line to a set of data points. The model can be written as:
y = w0 + w1 x

Given training data:


x y
1.0 2.0
2.0 2.8
3.0 3.6
4.0 4.4

The initial weights are w0 = 0.5 and w1 = 0.5, and the learning rate is α = 0.01. The loss function is the Mean
Squared Error (MSE):

a) Compute the gradient of the loss function with respect to w0 and w1 .


b) Using Gradient Descent, update the weights w0 and w1 after one iteration. Show all calculations.

4. Define Decision Trees. Discuss their structure and how they make decisions based on input features.
Consider the following dataset used for classifying whether a customer will buy a product based on their age,
income, and marital status. The training data is as follows:
Age Income Marital Status Bought Product (Target)
25 50000 Single No
30 60000 Single Yes
35 80000 Married Yes
40 90000 Married Yes
45 100000 Single No
a) Compute the Gini Impurity for the entire dataset. Show all calculations.
b) Based on the provided data, calculate the Gini Impurity for a potential split on Marital Status (Single vs.
Married) and determine if this is a good split. Show all calculations.

You might also like