Scribd
Upload
2 views2 pages

Assignment 1 Simple Linear Regression

The assignment requires implementing a simple linear regression model using gradient descent on provided datasets. Students must answer specific questions regarding the model's performance, learning rates, and cost functions, while adhering to guidelines on data normalization and code implementation without certain libraries. The final submission must include a PDF report with answers and plots, along with a GitHub link to the code.

Uploaded by

s45210488
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
Download as docx, pdf, or txt
2 views2 pages

Assignment 1 Simple Linear Regression

The assignment requires implementing a simple linear regression model using gradient descent on provided datasets. Students must answer specific questions regarding the model's performance, learning rates, and cost functions, while adhering to guidelines on data normalization and code implementation without certain libraries. The final submission must include a PDF report with answers and plots, along with a GitHub link to the code.

Uploaded by

s45210488
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1/ 2

Assignment 1

Simple Linear Regression

In this assignment, you need to implement a simple linear


regression model using gradient descent algorithm on the
following database.

Database
Independent/Predictor Variable -
https://drive.google.com/file/d/1cFZHElm5ebn1OolgmrLOxw91T8am3
S5C/view?usp=sharing

Dependent/Response Variable -
https://drive.google.com/file/d/
1rY3oTHxa1FT3fdcjWuKKQ4DOXnU8AeQM/view?usp=sharing

You are required to use the following Cost function

Answer the following questions based on your observations


1. Use linear regression to fit a straight line to the given
database. Set your learning rate to 0.5. What are the
cost function value and learning parameters values after
convergence? Also, mention the convergence criteria you
used.

2. The cost function that we are using in this assignment is


different than the one we used in class. Can you think of
the advantage of averaging the cost?

3. Plot cost function v/s iteration graph for the model in


question 1 for first 50 iterations.
4. Plot the given dataset on a graph and also print the
straight line you obtained in question 1 to show how it
fits the data.

5. Test your regression model with the learning rates


lr = 0.005, lr = 0.5, lr = 5
For each learning rate, plot a graph showing how the cost
function changes for the first 50 iterations and write
your observation.

6. Choose a suitable learning rate, then implement


stochastic and min-batch gradient descent, plot the cost
function against iteration, and observe how your cost
function changes compared to batch gradient descent.

Note

1. Use batch gradient descent unless mentioned in the


question.
2. Your code should be in Python.
3. Normalize the data before training.
4. You are required to implement the linear regression model
yourself. You are not allowed to use libraries like sci-
kit learn, Tensorflow, Pytorch, etc, and use their built-
in linear regression models.
5. You may use libraries like numpy, pandas, matplotlib, etc
to read and manipulate the dataset and plot the graphs.
6. Your plots should be labeled properly.
7. Normalize your predictor before starting training the
model.

How to Submit?
1. You need a submit a single PDF report file and the name
of that file should be your roll numbers.
2. Your report should have answers to all the questions with
the plots. Upload your code on github and mention the
link in the report. Do not copy paste code in the report.

You might also like