Udemy – Clustering And Dimensionality Reduction – Deep Dive 2024-1

Published on: 2024-08-08 19:55:29

Categories: 28

Descriptions

Clustering And Dimensionality Reduction – Deep Dive, Unsupervised learning is immensely important because it allows us to find hidden patterns and structures in data (a process also known as data mining) without the need for pre-labeled examples. This approach is not just useful, but often essential in situations where labeling data is impractical or impossible. In this course, our focus will be on two of the most impactful techniques in unsupervised learning: cluster analysis and dimensionality reduction. Clustering helps us to group similar data points based on their characteristics, uncovering underlying patterns in a dataset. Dimensionality reduction, on the other hand, simplifies complex data sets, making them easier to work with and understand. Mastering these techniques is key to deriving crucial insights from data which is a vital skill in the field of data science.

Cluster analysis and dimensionality reduction have widespread applications in data mining across various sectors. In marketing, they enable deeper customer insights and market segmentation. Healthcare professionals utilize them for analyzing patient data and identifying patterns in diseases. In the financial sector, these techniques are crucial for risk analysis and detecting fraudulent activities. They are also used in bioinformatics for interpreting genetic information. In e-commerce, these methods enhance product recommendation systems, while in social network analysis, they aid in understanding community patterns. Additionally, they’re applied in urban planning for traffic analysis. Beyond these, there are numerous other applications across wide range of industries. The aim of this course is in depth analysis of unsupervised learning algorithms. We’ll dissect these algorithms, explaining their inner workings, best practices, and limitations. This deep understanding is achieved not just through theory, but also by implementing the algorithms ourselves.

What you’ll learn

Unsupervised learning & data mining in python
Cluster analysis and dimensionality reduction
K-means based clustering (k-means, k-modes, k-prototypes)
Hierarchical (agglomerative clustering)
Agglomerative clustering linkages: Min, Max, Average and Wald
Density based clustering (DBSCAN, HDBSCAN)
Density based clustering validation (DBCV)
Graph based clustering (Louvain algorithm)
PCA dimensionality reduction
UMAP dimensionality reduction
Algorithms pros & cons
General guidelines for algorithms
Multiple approaches for preprocessing data for cluster analysis & dimensionality reduction
Metrics for cluster quality analysis
Comparing data clusterings

Who this course is for

Beginner/aspiring data professionals wanting to learn about unsupervised learning & data mining.
Intermediate/advanced data professionals wanting to improve their knowledge of unsupervised learning & data mining.

Specificatoin of Clustering And Dimensionality Reduction – Deep Dive

Publisher : Udemy
Teacher : Dalibor Veljkovic , Soledad Galli
Language : English
Level : All Levels
Number of Course : 203
Duration : 27 hours and 40 minutes

Content of Clustering And Dimensionality Reduction – Deep Dive

Clustering And Dimensionality Reduction - Deep Dive

Requirements

Knowing how to install python, python packages and set up jupyter notebook.
Basic python knowledge and the ability to execute python code in a jupyter notebook.
Knowledge of fundamental matrix operations: addition, multiplication, transpose, …
Understanding basic data analysis concepts such as mean, standard deviation or median.
Understanding basic math functions (e.g. square root, logarithms).
Basic understanding of the derivatives of elementary functions and the application of the chain rule in simple scenarios. This knowledge is only required for the UMAP chapter.