0% found this document useful (0 votes)
4 views11 pages

Deep Learning Algorithms Explained

Uploaded by

Aryan Kamath
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
4 views11 pages

Deep Learning Algorithms Explained

Uploaded by

Aryan Kamath
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Rolling in the

Deep...Learning
Algorithms

“We could’ve had it all. But you forgot to preprocess your data.”
- Adele (Supposedly)

Aryan Kamath
With a Little Help From My Google
Search Bar
While My GPU Gently Weeps
While My GPU Gently Weeps (cont.)
NN Nodes Forever

◉ Deep Learning may seem intimidating (and it is), but


it’s just a Neural Network with MANY more nodes

◉ Made possible by the advent of cloud storage of large


datasets and increased computational power

◉ Many different types of algorithms:


○ Multilayer Perceptron (ie. a Neural Network with many layers)
○ Convolutional Neural Network
○ Recurrent Neural Network
Lucy in the Network with
Handwritten Numbers

◉ We’re gonna look at an example of a Convolutional Neural


Network used to identify handwritten numbers

◉ Before running an image through a Neural Network, you do


multiple “convolutions”

◉ The point of a convolution is to find features that a normal


Neural Network would have trouble finding

◉ Then run the convolved images in your regular Neural


Network
Lucy in the Network with
Handwritten Numbers (cont.)

◉ We’re gonna look at an example of a Convolutional Neural


Network used to identify handwritten numbers
Lucy in the CNN with Handwritten
Numbers (cont.)

◉ This is what filters would start like


An Epoch in the Life

◉ This is what filters would end like


Here Comes the Code

◉ [Link]
Z424uv0iUe-LrCglL51o_WDa6E9#scrollTo=VLkit
CvTOGjt
Citation Lane/Tomorrow Never
Knows

Honglak Lee, Roger Grosse, Rajesh Ranganath, and Andrew Y. Ng. 2009.
Convolutional deep belief networks for scalable unsupervised learning of
hierarchical representations.

Ren, S., He, K., Girshick, R., and Sun, J. 2015. Adam: A Method for Stochastic
Optimization

Further research into this topic would include a deeper discussion on


activation functions, loss functions, optimizers, attention and other
types of algorithms for different datasets specifically text based ie.
GPT3.

3Blue1Brown’s Youtube Series explaining Neural Networks with visuals


[Link]
00Dx_ZCJB-3pi

You might also like