Dimension reduction: from modeling to visualization


As developers, engineers, scientists or students if we analyze data or want to learn how to analyze data, we need to have the required skills. We need to learn those skills helping us to improve our status in our current job, get a better job and be successful in next interviews we will have. Dimension reduction is among the subjects necessary to know if you want to be successful or become successful as developers, scientists or engineers.

In this workshop, you will learn about widely used dimension reduction methods such as PCA, ICA, t-SNE, and UMAP. You will learn how to explore them to reduce dimensionality of your data in Python, how to use methods like PCA in supervised learning schemes and how to improve your reports using visualization techniques like t- SNE. This will be a hands on workshop in which we will work on multiple datasets and will learn how to implement each one of the aforementioned methods on them in Python

  • Session 1: Mar. 5th Tue 10am-12pm PST
  • Session 2: Mar. 7th Thu 10am-12pm PST

  • 8 Topics / 4 hours live talks
  • 3 hands-on code labs
  • Real time interactions with instructors
  • Watch sessions replay, disucssions, course notes any time

Check the content tab for full course outlines.

Dimensionality reduction is a mandatory skill for data analysis. The materials are prepared in a way that both beginners and experts working with data in any level will learn how to use each one of the dimensionality reduction methods.

Reasonable background in programming with python

The course live sessions have ended, you can still enroll to learn the course with recorded videos, slides, interactive code labs on jupter notebook, course notes and discussions.
Not refundable.

Module 1: Introduction to dimensionality reduction
  • Feature selection versus feature extraction/construction
  • Preserving global versus local structure of data
  • Reducing dimensionality of multiple datasets using PCA and ICA
  • Code lab 1 (PCA versus ICA for dimensionality reduction)

Module 2: Dimensionality reduction for visualization and classification
  • Using t-SNE to preserve local structure of data
  • Correct interpretation of t-SNE plots
  • UMAP versus t-SNE
  • Code lab 2 (t-SNE and UMAP for visualization)
  • Use new dimensions of some of these methods in supervised learning models
  • Code lab 3 (Using reduced dimensions for building classification models)
Ali Madani

Ali is currently a Ph.D. candidate at the University of Toronto working on the development of new machine learning models to predict cancer patients responses to drugs. During his study, he published research papers in high impact scientific journals and international conferences covering such fields as ensemble learning, unsupervised clustering. He has earned a master of a mathematics degree, focusing on modeling of stochastic processes in complex biological systems from the University of Waterloo. His linkedin: link
  • Start Date: On-demand
  • Venue: Online
  • Fee:
    $59 $9
  • Students enrolled:219
  • Status: learn on-demand
  • Preview this course:

Enroll This Course