Description:
Cluster analysis is a vital component of unsupervised learning and data science. In this short session, we will explore the couple of unsupervised learning techniques – perform data clustering, reduce data dimensionality.
Setup:
Because this is an abbreviated session, attendees MUST install Anaconda software https://www.anaconda.com/ and have a basic understanding of using Jupyter Notebook.
Course Code/Duration:
BDT114 / 90 Minutes
Learning Objectives:
We will learn the following about the recommendation systems:
- Build a model to create clusters from data. Understand the intuition behind what principal component analysis (PCA).
- Build a simple PCA model to reduce dimensionality of data
- This session is designed for anyone who is familiar with basic steps involved in machine learning and are familiar with tools involved in building machine learning models.
Basic understanding of python language, pandas library and understanding of how to use Juypter Notebook.
Course Outline:
Introduction to Unsupervised Learning
- Definition and key concepts of unsupervised learning in machine learning.
Clustering Algorithms
- Overview of popular clustering methods (e.g., K-means, Hierarchical clustering) and their applications.
III. Dimensionality Reduction
- Techniques such as Principal Component Analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE) for reducing the dimensionality of data.
Real-world Applications and Best Practices
- Practical examples of unsupervised learning applications, along with best practices for model evaluation and interpretation.
Training material provided: Yes (Digital format)
