ECE 695E Lecture 10: Big Data Classification by Principal Component Analysis

By Muhammad A. Alam

Electrical and Computer Engineering, Purdue University, West Lafayette, IN

View Presentation

Category

Online Presentations

Published on

Abstract

Outline

  • Introduction
  • Why do we need reduction in data dimension
  • Theory of Principle Component Analysis
  • Applications of Principle Component Analysis
  • Conclusions

Cite this work

Researchers should cite this work as follows:

  • Muhammad A. Alam (2019), "ECE 695E Lecture 10: Big Data Classification by Principal Component Analysis," https://nanohub.org/resources/29419.

    BibTex | EndNote

Time

Location

2279 Wang, Purdue University, West Lafayette, IN

ECE 695E Lecture 10: Big Data Classification by Principal Component Analysis
  • Lecture 10. Big Data Classification by Principal Component Analysis 1. Lecture 10. Big Data Classific… 0
    00:00/00:00
  • copyright 2018 2. copyright 2018 71.604938271604937
    00:00/00:00
  • Course Outline 3. Course Outline 72.005338672005337
    00:00/00:00
  • Big vs. small data 4. Big vs. small data 73.773773773773783
    00:00/00:00
  • Small vs. big data 5. Small vs. big data 129.46279612946279
    00:00/00:00
  • Where do data come from? 6. Where do data come from? 217.81781781781783
    00:00/00:00
  • … driven by memory technology 7. … driven by memory technolog… 225.85919252585921
    00:00/00:00
  • 8. "Big data" techniques apply to… 281.8818818818819
    00:00/00:00
  • Our goal for the next few lectures … 9. Our goal for the next few lect… 334.60126793460125
    00:00/00:00
  • Analysis of big data 10. Analysis of big data 458.82549215882551
    00:00/00:00
  • Outline 11. Outline 575.90924257590927
    00:00/00:00
  • Classification problem in big data 12. Classification problem in big … 578.91224557891223
    00:00/00:00
  • PCA helps classification 13. PCA helps classification 764.16416416416416
    00:00/00:00
  • PCA Also help in data compression 14. PCA Also help in data compress… 946.88021354688021
    00:00/00:00
  • Outline 15. Outline 1027.1604938271605
    00:00/00:00
  • Principle Component Analysis (PCA) 16. Principle Component Analysis (… 1027.4607941274608
    00:00/00:00
  • Basic Concept of PCA 17. Basic Concept of PCA 1111.7784451117784
    00:00/00:00
  • PCA through Singular Value Decomposition 18. PCA through Singular Value Dec… 1279.5462128795464
    00:00/00:00
  • Reduce dimension by Singular Value Decomposition 19. Reduce dimension by Singular V… 1416.5832499165833
    00:00/00:00
  • Example 1: Rotation matrix 20. Example 1: Rotation matrix 1422.7560894227561
    00:00/00:00
  • SVD rotates the axes optimally 21. SVD rotates the axes optimally 1514.5145145145145
    00:00/00:00
  • SVD components allows reconstruction 22. SVD components allows reconstr… 1690.4571237904572
    00:00/00:00
  • Projection along PCs 23. Projection along PCs 1824.4911578244912
    00:00/00:00
  • Example 2: More general result 24. Example 2: More general result 1923.6903570236905
    00:00/00:00
  • SVD approximates the exact result 25. SVD approximates the exact res… 2026.9269269269271
    00:00/00:00
  • (continued) Projection along PCs 26. (continued) Projection along P… 2095.1618284951619
    00:00/00:00
  • Outline 27. Outline 2178.511845178512
    00:00/00:00
  • Principle Component Analysis for classification 28. Principle Component Analysis f… 2184.9849849849852
    00:00/00:00
  • Image Transmission by Principle Component Analysis 29. Image Transmission by Principl… 2294.5278611945278
    00:00/00:00
  • MATLAB code (by Camsari) 30. MATLAB code (by Camsari) 2321.7217217217217
    00:00/00:00
  • Conclusions 31. Conclusions 2394.0607273940609
    00:00/00:00
  • Review Questions 32. Review Questions 2576.2429095762432
    00:00/00:00
  • References 33. References 2835.4020687354023
    00:00/00:00
Pop Out