0% found this document useful (0 votes)
9 views13 pages

Gradient Descent Speech New

Uploaded by

Alex Ambrioso
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as KEY, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
9 views13 pages

Gradient Descent Speech New

Uploaded by

Alex Ambrioso
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as KEY, PDF, TXT or read online on Scribd
You are on page 1/ 13

The mathematics of machine learning

How the Blippity-Blip Do


Computers Learn?

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


Warning this talk is rated
MC!!!

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


Purpose

To present a mathematical idea


fundamental to most computer learning
in an approachable way.

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISERS ON SEPTEMBER 7TH 2020


Scope of my talk….

Supervised Learning
Gradient Descent

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


Examples of Supervised
Learning
: Google
Translation

Imagen : fGoogle Research


DALL*E 2: OpenAI

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


What is a gradient?
Gradient descent: an
analogy

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


How can gradient descent
solve a learning problem?

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


What is a gradient?
Gradient descent: the math

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


An application: finding a
trend line

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


Defining the loss of a particular trend line….
The last piece: a loss
function

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


Demo of the solution

Code

Visualization

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


How does this work for other
tasks?
Everything is mathematical
Text-to-image task

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISER ON SEPTEMBER 7TH 2020


Bibliography
Gradient descent: https://en.wikipedia.org/wiki/Gradient_descent
Machine Learning: https://en.wikipedia.org/wiki/Machine_learning
Supervised learning: https://en.wikipedia.org/wiki/Supervised_learning

Loss function: https://en.wikipedia.org/wiki/Loss_function


But what is a neural network: https://youtu.be/aircAruvnKk

Imagen: https://imagen.research.google/
DALLE 2: https://openai.com/dall-e-2/

A simple Neural Network: https://github.com/aambrioso1/presentations/blob/master/programming_hcc_2022/simple_nn.py

Neural network output: https://github.com/aambrioso1/HCC-Programming-Project/blob/master/nn_output.png

Python: python.org

My website: alexambrioso.com

PRESENTATION BY ALEX AMBRIOSO FOR HIGHRISERS ON SEPTEMBER 7TH 2020

You might also like