Scribd
Upload
2 views

Day 4

The document outlines various applications of Artificial Intelligence and Machine Learning, focusing on linear regression, image compression, logistic regression, neural networks, and autoencoders. It explains how linear regression can be used for image compression through Discrete Cosine Transforms, and describes the functionality of autoencoders in unsupervised learning for tasks like dimensionality reduction and outlier detection. Interactive tutorials and playgrounds for practical experimentation are also provided.

Uploaded by

shathaalharbi14255
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
2 views

Day 4

The document outlines various applications of Artificial Intelligence and Machine Learning, focusing on linear regression, image compression, logistic regression, neural networks, and autoencoders. It explains how linear regression can be used for image compression through Discrete Cosine Transforms, and describes the functionality of autoencoders in unsupervised learning for tasks like dimensionality reduction and outlier detection. Interactive tutorials and playgrounds for practical experimentation are also provided.

Uploaded by

shathaalharbi14255
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 25

Artificial Intelligence and Machine Learning

Applications
Lecture Outline
• Linear Regression Tutorial & Playground
• Image Compression
• Logistic Regression Interactive Demo
• Neural Network Regression Tutorial & Playground
• Auto-Encoders
Linear Regression Tutorial

https://mlu-explain.github.io/linear-regression/
Linear Regression Playground

https://observablehq.com/@yizhe-ang/interactive-visualization-of-
linear-regression
Image Compression with Linear Regression
• An image can be transformed into the frequency domain and represented as a
combination of some basic components.
• Cosine Basis Functions and Discrete Cosine Transforms (DCT) can enable this.
• The human eye is most sensitive to low frequencies. Therefore, most of the high
frequencies can be ignored.
• Principal behind JPEG Image Compression.

https://bengarney.com/2016/06/25/video-conference-part-2-joint-photographic-hell-for-beginners/
Image Compression with Linear Regression
• Using linear regression, we can learn the weight of each of these
cosine basis.
• We only need to store the weights of the basis frequencies.
• The image is reconstructed using these weights.
Image Compression
8 x 8 Pixels Image

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Image Compression
• Gray-Scale Example:
• Value Range 0 (black) --- 255 (white)

63 33 36 28 63 81 86 98
27 18 17 11 22 48 104 108
72 52 28 15 17 16 47 77
132 100 56 19 10 9 21 55
187 186 166 88 13 34 43 51
184 203 199 177 82 44 97 73
211 214 208 198 134 52 78 83
211 210 203 191 133 79 74 86

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Image Compression
• 2D-DCT of matrix
Numbers are coefficients
of polynomial

-304 210 104 -69 10 20 -12 7


-327 -260 67 70 -10 -15 21 8
93 -84 -66 16 24 -2 -5 9
89 33 -19 -20 -26 21 -3 0
-9 42 18 27 -7 -17 29 -7
-5 15 -10 17 32 -15 -4 7
10 3 -12 -1 2 3 -2 -3
12 30 0 -3 -3 -6 12 -1

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Image Compression
• Cut the least significant components

-304 210 104 -69 10 20 -12 0


-327 -260 67 70 -10 -15 0 0
93 -84 -66 16 24 0 0 0
89 33 -19 -20 0 0 0 0
-9 42 18 0 0 0 0 0
-5 15 0 0 0 0 0 0
10 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0

As you can see, we save a little over half the original memory.

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Reconstructing the Image
• New Matrix and Compressed Image

55 41 27 39 56 69 92 106
35 22 7 16 35 59 88 101
65 49 21 5 6 28 62 73
130 114 75 28 -7 -1 33 46
180 175 148 95 33 16 45 59
200 206 203 165 92 55 71 82
205 207 214 193 121 70 75 83
214 205 209 196 129 75 78 85

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Can You Tell the Difference?
Original Compressed

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Image Compression
Original Compressed

Slide based on Discrete Cosine Transform and Image Compression by Thomas Buck et. Al.
Logistic Regression Interactive Tutorial

https://mlu-explain.github.io/logistic-regression/
Neural Networks Tutorial

https://mlu-explain.github.io/neural-networks/
Neural Networks Playground

https://playground.tensorflow.org/
Try it Yourself – Digit Classification

https://trekhleb.dev/machine-learning-
experiments/#/experiments/DigitsRecognitionMLP
Try it Yourself – Sketch Recognition

https://trekhleb.dev/machine-learning-
experiments/#/experiments/SketchRecognitionMLP
Applications – AutoEncoders
• Autoencoders are a type of neural networks where the input is also
the output.
• They come under unsupervised learning and there are no labels
involved.
• An autoencoder consists of two parts: encoder and decoder.
• The idea here is that you take a higher dimensional input, project it
into a lower dimensional space and then project it back into the input
space.
Applications – AutoEncoders

https://github.com/JamesAllingham/LaTeX-TikZ-Diagrams
Applications – AutoEncoders
• The autoencoder model tries to minimize the reconstruction error
(RE).
• Typically, mean squared is used as the loss function for autoencoders.
• The objective is to minimize the following:
AutoEncoders for Dimensionality Reduction
• The encoder output can be used for dimensionality reduction

https://towardsdatascience.com/dimensionality-reduction-with-autoencoders-versus-pca-f47666f80743
AutoEncoders for Outlier Detection
• The reconstruction error can be used to detect outliers. Out of distribution
samples will have high Reconstruction error.

Outliers

https://towardsdatascience.com/dimensionality-reduction-with-autoencoders-versus-pca-f47666f80743
AutoEncoders for Image Completion
Corrupted Corrupted
Original Input AE Output Original Input AE Output

Smieja et. al. Can auto-encoders help with filling missing data?
Summary

You might also like