Scribd
Upload
15 views2 pages

ML Assignment 2

ML

Uploaded by

Nishant Sharma
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
15 views2 pages

ML Assignment 2

ML

Uploaded by

Nishant Sharma
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
You are on page 1/ 2

Maharaja Agrasen Institute of Technology, Delhi

Department of Information Technology


Assignment -2
Subject Name: Machine Learning
Subject Code: CIE-407T
Class: B.Tech (IT); Group-7AIML-IV-ABC, 7AIML-VABC
Faculty: Dr. Bhoomi Gupta, Dr. K.C.Tripathi

Last Date to submit: 26-09-2024

Questions:

Unit 1: Introduction to Machine Learning & Basic Concepts (CO1)

Question 1 - The "Machine Learning Needed" Pitch: Imagine you are pitching a new product
or service that heavily relies on machine learning to a panel of investors. These investors have
never heard of machine learning before. Your task is to:
● Explain why machine learning is essential for your product/service to succeed.
● Use clear, concise, and engaging language that avoids technical jargon.
● Connect machine learning to solving a real-world problem in a way that anyone can
understand.
● What are some common misconceptions about what machine learning can realistically
do? Where have you seen it overhyped?

Question 2 - The Data Dilemmas: You are a data scientist tasked with building a machine
learning model to predict customer churn for a social media company. However, you encounter
several data-related challenges:
● Missing Data: A significant portion of user engagement data is missing. How do you
handle this? What are the implications of different approaches?
● Biased Data: You discover the existing data reflects a bias toward a specific demographic.
How does this bias impact the model's fairness and accuracy, and how can you mitigate
it?
● High-Dimensional Data: The dataset contains hundreds of features about each user.
Explain the challenges of working with such data and propose dimensionality reduction
techniques like PCA, LDA, or ICA, highlighting their pros and cons.
Unit 2: Regression Analysis in Machine Learning (CO2)

Question 1 - The Case of the Confusing Coefficients: You have built a multiple linear
regression model to predict housing prices. To your surprise, some of the coefficient signs don't
make intuitive sense (e.g., the number of bedrooms has a negative coefficient).

● Explain why this situation might occur, referring to concepts like multicollinearity and
confounding variables.
● What if our data was NOT normally distributed? What if we had missing values? How
would this change our approach?

Question 2 - The "Beyond R-squared" Challenge: Many people mistakenly believe that a high
R-squared value always indicates a good regression model.

● Debunk this myth by explaining the limitations of R-squared as the sole evaluation
metric.
● Introduce alternative metrics such as adjusted R-squared, RMSE, and MAE, discussing
their significance and when to use them.
● Which value helps us know if the RELATIONSHIP between variables is statistically
significant? What does that REALLY mean?
● Emphasize the importance of considering the model's interpretability, generalizability,
and the specific goals of the analysis alongside statistical metrics.

Question 3 - The Logistic Regression Predicament: You are tasked with building a model to
predict customer churn (whether a customer will leave or not) for a telecommunications
company using logistic regression.
● Explain how logistic regression, despite its name, is suitable for this classification
problem.
● Discuss the key assumptions of logistic regression and how you would assess whether
they hold true for your data.
● Describe how you would interpret the output of the logistic regression model, focusing
on probabilities and odds ratios.

Note: 1. Write your name and roll number on the front page in clear writing.
2. Failure to submit in time will incur penalty in marks.

You might also like