DCA3106 MACHINE LEARNING

Description

SESSION	JULY 2026
PROGRAM	BCA
SEMESTER	V
COURSE CODE & NAME	DCA3106 MACHINE LEARNING

 

 

Set-I

 

Q.1. Describe the Machine Learning workflow. (10 Marks)

Ans 1.

Meaning of the Machine Learning Workflow

The machine learning workflow is a series of steps that a developer takes to get from raw data to a working model. It’s not a one-step process; missing or skipping any of the steps will most likely result in a weak or unreliable model.

Gathering and Analyzing Information

The first step is to collect data from a relevant sources like databases, sensors, surveys and existing files. After it’s collected

Its Half solved only

Buy Complete from our online store

 

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

 

MUJ Fully solved assignment available for session Jul-Aug 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.

 

 

Q.2. Differentiate between supervised learning and unsupervised learning with suitable example. (10 Marks)

Ans 2.

Meaning of Supervised Learning

Supervised learning involves an algorithm that learns about data, for which the right answers are already known. In each training example, the input is presented together with the desired output, called a label, and the goal of the model is to discover the mapping that relates the input to the output and then make predictions for new, unlabeled inputs. The prediction of house prices, as a function of area, location, number of rooms etc., is a good example because the selling prices of houses in the training data are known.

Meaning of Unsupervised Learning

Unsupervised learning is applied to data without any labels. It is just fed a set of inputs and left to itself to determine the structure, patterns, or groupings. A common example is customer segmentation, where a retail company segments customers by their purchasing habits without a

 

 

Q.3. Discuss the need for feature scaling in KNN and explain normalisation and standardization techniques. (10 Marks)

Ans 3.

Role of Feature Scaling in KNN

The K Nearest Neighbours algorithm takes a new data point and computes its distance from the other points in the data set, then examines the labels of the other data points that are the closest. This implies that the complete algorithm relies on distance (typically Euclidean distance) computations between feature values. This distance calculation becomes distorted if the features in a given dataset have measurements that are on very different scales.

Problem Caused by Unscaled Features

Suppose you are having

 

 

Set-II

 

Q.4. List and explain the different types of Naïve Bayes classifiers. (10 Marks)

Ans 4.

Basic Idea Behind Naive Bayes

Naive Bayes is a classification method that is based on Bayes theorem, which determines the probability of a specific outcome based on a set of evidence. It is termed “naive” because it assumes all features in the dataset are independent, and in the real data, this assumption is not true, but the algorithm still works well in practice. Based on the nature of data to be processed, various variants of Naive Bayes are used.

Gaussian

 

 

Q.5. What are support vectors? Why are they important. (10 Marks)

Ans 5.

Meaning of Support Vectors

The Support Vector Machine algorithm’s key idea is to identify the optimal hyperplane that separates data points from different classes. These are the data points in the training set that lie closest to this separating hyperplane and are known as support vectors. These points are not located far into their class regions, but just on the edge of each region, in the middle of the boundary lines.

Role of

 

 

 

Q.6. Discuss the need and importance of cluster analysis in modern data mining and machine learning applications. (10 Marks)

Ans 6.

Meaning of Cluster Analysis

Unsupervised machine learning method that categorizes a collection of data points into clusters so that data points in a cluster are more alike than those in different clusters. Because there are no labels in the data that is incorporated into the algorithm, the algorithm must uncover these natural groups based on patterns in the features themselves, not on any idea of what the grouping should be.

Need for Cluster Analysis in Data Mining

The data gathered today, from retail sales, social media interactions, sensor networks, and much more, is often too large to have categories preattached to it. Analysts need to make sense of this raw, unlabelled data, and to do so they need to find structure in it that they would otherwise not be able to see in thousands or millions of individual records: hence the need for cluster analysis. Otherwise, there would be no way to manually search for patterns in such huge datasets.

Importance