Clustering in Machine Learning

Name: Clustering in Machine Learning
Author: Briana Brownell

Learn the basics of machine learning and its applications.

Preview the first lesson free — get full access to all 37 lessons.

Course: On-Demand

Intermediate Provider : Briana Brownell 8 chapters · 37 Lessons · 2h 39m in Arabic, German, English, Spanish, French, Portuguese, Urdu, Chinese

Course Description

In this course, you’ll learn the basics of a variety of areas of machine learning, including clustering, anomaly detection, and association modeling. You’ll understand the most common method of clustering, k-means, and how it can be applied to survey data in order to create a robust customer segmentation, including the creation of personas, and a typing tool. By the end of this course, you’ll know how to perform a number of different clustering and analysis techniques.

More specifically, you’ll review the most common methods of clustering, along with their performance on several example datasets. You will learn how to analyze data for intentional and unintentional anomalies like fraud and data-entry errors, detect anomalous data, and see how to validate them. You’ll examine how to create association rules between purchases and how they can inform the strategy of a retail store or an e-commerce website. Common challenges will be covered so that you can ensure the models you create are valid and robust. Finally, you’ll have the chance to review several case studies so you can see how what you’ve learned about work in detail.

What You'll Learn

Understand the various clustering algorithms and their strengths and weaknesses, including k-means, mini-batch k-means, agglomerative, spectral, DBSCAN, and Gaussian mixture models
Choose the best clustering algorithm for a specific problem and identify where clustering applies to real-world problems
Apply k-means to survey data to build customer segmentation, personas, and a typing tool
Detect intentional and unintentional anomalies such as fraud and data-entry errors using a variety of methods and validate them
Build association rules using support, confidence, and lift to understand purchase behavior
Analyze case studies of clustering and anomaly-detection models that have been implemented

Key Takeaways

K-means is the most common method of clustering and can be applied to survey data to create customer segmentation, personas, and a typing tool.
Clustering covers a range of unsupervised techniques and selecting the right algorithm depends on the specific problem and dataset.
Anomaly detection can find both intentional anomalies like fraud and unintentional ones like data-entry errors, and detected anomalies should be validated.
Association rules describe relationships between purchases and can inform the strategy of a retail store or an e-commerce website.
Addressing common challenges helps ensure the clustering and analysis models created are valid and robust.

Frequently Asked Questions

What topics does this course cover?

The course covers the basics of several areas of machine learning, including clustering, anomaly detection, and association modeling. It reviews common clustering methods and their performance on example datasets, anomaly detection, building association rules, and case studies of implemented models.

Which clustering algorithms will I learn?

You'll review clustering methods including the k-means algorithm, mini-batch k-means, agglomerative clustering, spectral clustering, DBSCAN, and Gaussian mixture models, along with an algorithm comparison.

What skills will I gain from this course?

You'll gain skills in cluster analysis, data science, hierarchical clustering, k-means clustering, machine learning methods, and spectral clustering.

How is k-means applied in this course?

K-means is applied to survey data to create a robust customer segmentation, including the creation of personas and a typing tool, and the course also covers choosing input variables, choosing the number of clusters, centroid interpretation, and model-quality metrics.

What will I be able to do with association rules?

You'll learn how to create association rules between purchases using support, confidence, and lift, and see how they can inform the strategy of a retail store or an e-commerce website.