University of Westminster

Trends in computer Science

4COSC008C

Machine Learning
Overview of machine learning. Describe and compare two different machine
learning techniques.

Full Name : Malsha Pabodani Jayasinghe

UoW Number : w2051880
IIT Number : 20231109

Group Members

Full name UoW Number

Ammar Aflal w2051786

Deshitha Ganegoda w2051830
Shaheed Razzak w2051856

1
Table of Contents

1. INTRODUCTION……………………………....………….………………………………………....3

2. OVERVIEW OF MACHINE LEARNING………...………………………..………...4

2.1 What is Machine Learning? ……………………………………………………………….4

2.2 Significance of Machine Learning ……………………………………………………...4

3. MACHINE LEARNING TECHNIQUES…..………………………..………………...5

3.1 Supervised Learning………………………..…………………………………………………5

3.1.1 Classification………………………………………………………………………...5

3.1.2 Regression…………………………………………………………………………….6

3.2 Unsupervised Learning………………………………………………………………………6

3.2.1 Clustering……………………………………………………………………………...6

3.2.2 Association Rule Learning……………………………………………………..7

3.2.3 Dimensionality Reduction………………………………………...……………7

4. COMPARISON OF SUPERVISED LEARNING AND UNSUPERVISED

LEARNING…………………………………………………………..…………………..…………………8

5. CONCLUTION…...…..……………….……………………………………………………..…………….9

REFERENCES……………………………………………………………………………………..…………10

2
1. INTRODUCTION

With the development of technology, machine learning was developed by man to give
machines the ability to learn and make decisions by considering human learning methods. At
present, machine learning is used to improve the efficiency and productivity of machines. There
are several technical methods of machine learning.
This research aims to get an overview of machine learning and focuses on two types of machine
learning techniques.

3
2. OVERVIEW OF MACHINE LEARNING

Figure 1 – Machine Learning

Source: Adapted from [10]

2.1 What Is Machine Learning?

Machine learning is a process in which we can teach a machine to learn using data that we
are given. It can identify patterns using algorithms and statistical models using that data. In this
process, machines get the ability to learn from the identified patterns and predict solutions to our
problems in the future [5],[9].

2.2 Significance Of Machine Learning

A typical computer can only process data using the introduction given by a human. It cannot
predict its own instructions to make decisions.

In the case of machine learning, it can learn from past data and create instructions for its
own process. In this case, we do not need to give instructions at every point. The machine learning
process can give the ability to the machine to identify the problem and solve the problem itself
[8],[3].

4
3. MACHINE LEARNING TECHNIQUES

Figure 2- Machine Learning Techniques

Source: Adapted from [11]

3.1 Supervised Learning

Supervised learning is one of the machine learning techniques that uses labelled data for its
learning process. In this technique, the machine needs labelled data to train its models. The
machine identifies patterns using labelled data with supervised learning algorithms and models.
Then, it predicts solutions in future [5].

Supervised learning has two methods.

• Classification
• Regression

3.1.1 Classification

Classification is a method under supervised machine learning that is trained to predict

discrete values. By using past experiences, it predicts the output by classifying according to the
categories. Under classification tasks, there are several algorithms such as decision trees, rule-
based classifiers, Naïve-Bayesian classifiers, K-nearest neighbour classifiers, neural networks,
Linear Discriminant Analysis, and support vector machines. [1],[7],[9].
5
3.1.2 Regression

Another method of Supervised learning is Regression. Algorithms and statistical models

under Regression are usually used to predict outputs that have numerical values using labelled data
as past experiences. Regression approach is used when faced with problems like ,The shares of the
company are tomorrow [6].

3.2 Unsupervised Learning

Another method of machine learning is unsupervised learning. Unsupervised learning aims

to take an input dataset and identify essential groups or patterns among the data points or records.
As a result, unsupervised learning is also called descriptive training, and learning without
supervision is called knowledge or pattern discovery [1],[3].

There are three methods of unsupervised learning.

• Clustering
• Association rule learning
• Dimensionality reduction

3.2.1 Clustering

“Clustering algorithms - The algorithms that group data into clusters based on similarity” [5].

Clustering is one of the methods in Unsupervised learning. So clustering algorithms learn

using unlabelled data. This method is quite similar to the classification method in supervised
learning. The difference between these two methods is that the clustering method uses unlabelled
data. It can identify similar data and group those data according to the similarities [2].

Figure 3- Clustering
Source : Adapted from [12]

6
3.2.2 Association rule learning

The association rule method is also another method in unsupervised learning. In this
method, algorithms can identify the relationships between the data. Simply, this can analyze data,
make the rule between it, and predict the following outcome. Using this method, the machine can
predict the outcomes, like recommending another product for the customer by analyzing the
products chosen by the customer [4].

3.2.3 Dimensional reduction

“Dimensionality reduction algorithms - The algorithms that simplify our data and faithfully
describe it with fewer features” [5].

Dimensional reduction is another method in Unsupervised learning. In this method, reduce

some similar inputs. The purpose of this is to make the process of learning for the machine easy.
When millions of features give in to the learning process to analyze the inputs, it is quite
challenging and makes the process slow. In dimensional reduction, it can identify and merge the
most similar features into one feature. By doing this, it will retain much information [2],[3].

7
4. COMPARISON OF SUPERVISED LEARNING AND
UNSUPERVISED LEARNING

Supervised learning and Unsupervised learning techniques are used to train models using
their algorithms by analyzing the given data. However, these two techniques are used for different
kinds of data to learn and train their models. Supervised learning uses labelled data, while
unsupervised learning uses unlabelled data [1].

Supervised learning cannot identify the input if there is no given mapped labelled data. But in this
case, the Unsupervised technique can clarify the input raw data, train its models, and make the
predictions by itself [2],[1].

But in the Machine Learning process, these two types of learning techniques are used for different
purposes.

As an example,
Supervised learning -
The process of classifying Spam emails is predicated by supervised machine learning under the
classification method. Firstly, it needs to give examples of spam emails as labelled data, and then
classification algorithms learn from it and train its models to identify spam emails in future [5].

Unsupervised learning -
The Association rule learning method can predict the customer's preferred item by using the items
that the customer has selected. This process does not need any labelled data. Algorithms in
Unsupervised learning can analyze the raw data and make predictions [4].

8
5. CONCLUSION
Machines can learn from past experiences. By using machine learning algorithms, machines can
train their models and predict solutions for problems. In the case of problem-solving, we need to
use different kinds of machine-learning techniques according to the problem. Different kinds of
problems need to use different kinds of methods. So, it is essential to have a clear idea of the
differences between machine learning techniques.

9
REFERENCES

[1] S. Dutt, S. Chandramouli, and A. K. Das, Machine Learning. Pearson Education India, 2018.
Accessed: Feb. 10,
2024.[Online].Available:https://learning.oreilly.com/library/view/machinelearning/97893895881
32/xhtml/chapter001.xhtml#ch1_5

[2] H. Rhys, Machine Learning with R, the tidyverse, and mlr. Manning Publication, 2020.
Accessed: Feb. 10, 2024. [Online]. Available: https://learning.oreilly.com/library/view/machine-
learning-with/9781617296574/OEBPS/Text/kindle_split_012.html

[3] A. Géron, Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 3rd ed.
O’Reilly Media,Inc., 2022. Accessed: Feb. 10, 2024. [Online]. Available:
https://learning.oreilly.com/library/view/hands-on-machine-learning/9781098125967/

[4] J. Bell, Machine Learning, 2nd ed. Wiley, 2020. Accessed: Feb. 23, 2024. [Online].
Available: https://learning.oreilly.com/library/view/machine-learning-
2nd/9781119642145/c07.xhtml
[5] L. Serrano, Grokking Machine Learning. Manning Publication, 2021. Accessed: Feb. 18,
2024. [Online]. Available: https://learning.oreilly.com/library/view/grokking-machine-
learning/9781617295911/Text/01.xhtml

[6] M. A. El Mrabet, K. El Makkaoui, and A. Faize, “Supervised Machine Learning: A Survey,”.

2021.Accessed:
Feb.18, 2024. [Online]. Available:
https://ieeexplore.ieee.org/document/9641998/authors#authors

[7] R. N. Thomas and R. Gupta, “A Survey on Machine Learning Approaches and Its
Techniques,” 2020. Accessed: Feb.20.2024.[Online].Available:
https://ieeexplore.ieee.org/document/9087123

[8] A. Srivastava, S. Saini, and D. Gupta, “Comparison of Various Machine Learning Techniques
and Its Uses in Different Fields,” 2019. Accessed: Feb.20.2024 [Online]. Available:
https://ieeexplore.ieee.org/document/8822068

[9] B. Abdualgalil and S. Abraham, “Applications of Machine Learning Algorithms and

Performance Comparison: A Review,” IEEE, Apr. 2020. Accessed: Feb. 23, 2024. [Online].
Available: https://ieeexplore.ieee.org/document/9077759

10
[10] An Introduction to Machine Learning, Its Importance, Types, and Applications. 2022.
Accessed: Feb. 28, 2024. [Online]. Available:
https://www.google.com/url?sa=i&url=https%3A%2F%2Fwww.fsm.ac.in%2Fblog%2Fan-
introduction-to-machine-learning-its-importance-types-and-
applications%2F&psig=AOvVaw3MviJCK2GE2qcyqvV7kib7&ust=1709020402824000&sourc
e=images&cd=vfe&opi=89978449&ved=0CBMQjRxqFwoTCIi41KLDyIQDFQAAAAAdAAA
AABBQ
[11] N. TOMAR, Types of Machine Learning. 2023. Accessed: Feb. 26, 2024. [Online].
Available:
https://www.google.com/url?sa=i&url=https%3A%2F%2Fidiotdeveloper.com%2Ftypes-of-
machine-learning%2F&psig=AOvVaw3pTzRPECCvgr8fdVMaK-
Av&ust=1709110218375000&source=images&cd=vfe&opi=89978449&ved=0CBMQjRxqFwo
TCIj23veRy4QDFQAAAAAdAAAAABAI
[12] Include Help, Example of clustering in data sets. 2023. Accessed: Feb. 26, 2024. [Online].
Available: https://www.includehelp.com/basics/cluster-analysis-in-data-mining.aspx

11