Clustering and Unsupervised Learning

Clustering and Unsupervised Learning

Divyani Tirthyani, Sunil Kumar, Shally Vats
Copyright: © 2024 |Pages: 21
DOI: 10.4018/979-8-3693-3609-0.ch005
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Machine learning (ML) is an approach driven by data, wherein computers acquire knowledge from information without requiring human interference. Artificial intelligence (AI) and machine learning (ML) have made significant contributions across diverse research domains, leading to enhanced outcomes. Clustering is defined as a fundamental challenge in various data-driven fields, representing an unsupervised learning model. Unsupervised learning methods and algorithms encompass the Apriori algorithm, ECLAT algorithm, frequent pattern growth algorithm, k-means clustering, and principal components analysis. Unsupervised learning methods have achieved notable success in fields such as machine vision, speech recognition, the development of autonomous vehicles, and natural language processing. This chapter provides a brief explanation of unsupervised clustering approaches. It also discusses literature review, intriguing challenges, and future prospects in the realm of unsupervised deep clustering.
Chapter Preview
Top

1. Introduction

Clustering serves as a valuable asset in the toolkit of data science. It helps find patterns in data where things are similar within groups but different between groups. Biologists and social scientists first used hierarchical clustering. It's a part of statistical multivariate analysis. Clustering is unsupervised, meaning it doesn't need labelled data. There are two main types: those based on probability models and those without specific assumptions. Probability model-based approaches assume that data comes from a mix of probability models. They use a method called the expectation and maximisation (EM) algorithm to cluster the data. Nonparametric approaches use objective functions based on how similar or different data points are. These methods are often divided into hierarchical and partitional methods, with partitional methods being more commonly used (Sinaga & Yang, 2020). Unsupervised learning is important for making AI systems more like humans. That's because these systems need to figure things out on their own from lots of data without labels. Unsupervised learning is great for handling complicated tasks, unlike supervised learning, which is better at giving exact answers since programmers tell it what to learn from the data. Conversely, unsupervised learning can uncover surprises. It's what powers artificial neural networks, which are crucial for deep learning to happen. But even in neural networks, you can use supervised learning if you already know what results you want. Sometimes, learning without supervision is the aim itself. Unsupervised learning models, for example, can reveal hidden patterns in big sets of data and sort them into groups based on how similar or different they are (Naeem et al., 2023).

Below are several explanations highlighting the significance of unsupervised learning:

  • ●

    Having lots of data without labels is a big chance for learning.

  • ●

    Labelling data takes a lot of time and people.

  • ●

    Machine learning can make this easier for everyone involved.

  • ●

    Unsupervised learning is great for looking at new or messy data.

  • ●

    It's especially good for dealing with huge amounts of data and finding patterns.

1.1 Motivation

Here's why we're dedicating a chapter to clustering and unsupervised learning in healthcare:

  • ●

    A vast amount of healthcare information is being produced from diverse origins such as electronic health records, medical imaging, wearable technology, and genomic data. Using clustering and unsupervised learning helps make sense of all this data.

  • ●

    Personalised medicine is becoming more important. Clustering algorithms help doctors group patients based on their similarities, making it easier to create treatment plans that fit each person.

  • ●

    Healthcare organisations want to improve population health by identifying groups of people who are at risk for certain diseases. Clustering and unsupervised learning help with this, making it easier to target interventions and preventions.

  • ●

    Healthcare organisations need to use their resources wisely. Clustering methods can help them see patterns in things like patient admissions and procedures, so they can plan better and be more efficient.

  • ●

    Technology is getting better, making clustering and unsupervised learning more accessible and powerful. Using these techniques in healthcare can lead to new innovations and better care for patients.

  • ●

    Healthcare and data science are coming together, so it's important for doctors, data scientists, and researchers to work together. This chapter will give them a good overview of how clustering and unsupervised learning can improve patient care and research in healthcare.

Complete Chapter List

Search this Book:
Reset