Ben Simondet
University of Minnesota - Morris
April 30, 2016
Introduction
Background
Ways PCA Can Help
Conclusions
Introduction
Background
Methods
Conclusions
Introduction
Background
Methods
Conclusions
First Principal Component
Second Principal Component
Large eigenvalues = More influence
Large eigenvalues = More influence
Small eigenvalues
= Less influence
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
2-D to 1-D Graph
120-D (not possible as a physical graph)
to 2(or 3)-D Graph
Billions | Month | Eigenvalue |
---|---|---|
.66 | 11 | 2 |
X=11*2
Y=.66*2
(22,1.32)
Introduction
Background
Methods
Conclusions
Facial Recognition
Emotion Recognition
Eye Tracking
Application: Web Based Door Access
Algorithm Change
Algorithm Change
Facial Recognition
Emotion Recognition
Eye Tracking
Application: Assistance with Emotion Recognition
Application: Assistance with Emotion Recognition
Application: Assistance with Emotion Recognition
Facial Recognition
Emotion Recognition
Eye Tracking
Application: Assistance with ASD Diagnosis
Application: Assistance with ASD Diagnosis
Introduction
Background
Methods
Conclusions
Current Obstacles
What does the future hold?