DCA71M5 INTRODUCTION TO MACHINE LEARNING (ELECTIVE)

Description

SESSION	FEB-MARCH 2026
PROGRAM	MASTER OF COMPUTER APPLICATIONS (MCA)
SEMESTER	III
COURSE CODE & NAME	DCA71M5 INTRODUCTION TO MACHINE LEARNING (ELECTIVE)

 

 

 

Assignment Set – 1

 

 

Q1. A hospital wants to predict whether a patient will develop diabetes. Identify which type of Machine Learning should be used and justify your answer.

Ans 1.

Identifying the Type of Machine Learning

If a hospital wants to determine whether the patient is likely to develop diabetes or not, Supervised Learning is the best kind of machine learning that they can employ. Particularly, it is an issue of binary classification, which means that the model has to predict either or the patient is likely to develop the disease (positive classification) or the patient will not be diagnosed with the disease (negative classification). The best method is supervised learning. option because historical data that is labelled is readily available. This

 

Its Half solved only

Buy Complete from our online store

 

https://smuassignment.in/online-store/

 

MUJ Fully solved assignment available for session Jan-Feb 2026.

 

Lowest price guarantee with quality.

Charges INR 198 only per assignment. For more information you can get via mail or Whats app also

Mail id is aapkieducation@gmail.com

 

Our website www.smuassignment.in

After mail, we will reply you instant or maximum

1 hour.

Otherwise you can also contact on our

whatsapp no 8791490301.

 

 

 

Q2. An email system needs to classify emails as spam or not spam. Which supervised learning technique would you use and why?

Ans 2.

Problem Overview

The classification of emails as spam or not is an old binary classification issue in machine learning supervised. Email systems already have an extensive collection of email which were previously classified as spam or not by administrators or users. The fact that these historical emails have been labeled makes the decision to use supervised learning a natural one. It is the goal of training an algorithm on these labeled instances so that it is able to quickly and precisely classify the new email messages that are received.

Recommended Technique: Naive Bayes Classifier

The most popular and tested method for classifying spam is to use that of the Naive Bayes classifier, particularly the Multinomial Naive Bayes variant. The algorithm is based on Bayes theorem, and employs an approach based on probabilities to the classification of text. It determines the likelihood that an email message

 

 

Q3. In a cancer detection system, explain why Recall is more important than Accuracy. Support your answer with an example.

Ans 3.

Understanding Accuracy and Recall

Accuracy refers to the percentage of predictions total that are accurate. This can be misleading if the distribution of classes is not balanced which is often the case when it comes to cancer detection data in which healthy people outnumber cancer patients.

Recall, sometimes referred to as sensitive or the true positive rate, is the percentage of positive instances which the model is able to identify. The calculation is as follows: True Positives multiplied by (True Positives and False Negatives).

 

Assignment Set – 2

 

Q4. A company wants to analyze customer reviews from its website. Explain the preprocessing steps required before applying ML.

Ans 4.

Introduction

Reviews of customers are not structured information. Before any machine-learning algorithm is able to analyze sentiment, identify topics, or categorize opinions, the unstructured text has to be cleansed and converted into a structured numeral format. The process of transformation is known as text preprocessing. Preprocessing is a crucial factor in the accuracy and quality of the machine-learning model. The process of preprocessing for reviews of customers typically consists of

 

 

Q5. Explain how an OTT platform recommends movies to users using collaborative filtering. Compare content-based and hybrid recommendation systems using an example

Ans 5.

Collaborative Filtering on OTT Platforms

A OTT (Over-The-Top) service such as Netflix as well as Amazon Prime uses collaborative filtering to suggest movies to users. Collaboration-based filtering is based on the assumption that those who reached an agreement on preferences or ratings previously are more likely to be able to reach a consensus in the near future. The system does not rely on any details about the films

 

Q6. A company processes millions of transactions daily. Explain why traditional ML techniques may fail and how Big Data tools help.

Ans 6.

An organization that handles thousands of transactions every day generates an enormous, constant flow of information. Although traditional ML techniques are suitable for smaller to medium-sized data sets that can are able to fit into the memory of one machine however, they are severely limited in big datasets. Knowing these limitations and the ways big data-related tools can address these issues is crucial to build robust and reliable ML systems.

Why Traditional ML Techniques May Fail

Traditional ML algorithms such as Logistic Regression and Decision Trees as well as Support Vector Machines are usually developed to be run on a single computer. If the transaction data grows in the hundreds of gigabytes, or Terabytes, it is more than the storage and RAM capacity of a single server. The algorithm is unable to store the whole dataset in memory in order to train the model. This can result in out-of-memory errors or very slow