Chapitre 4

14/12/2021
Plan
Introduction
Régression
Apprentissage non supervisé
Arbres de décision
Apprentissage bayésien
Séparateurs a vaste marge
Réseaux de neurones artificiels
Modèles de Markov cachés
© FEZZA S. v21‐22
Apprentissage par renforcement
Machine Learning 1
Objectifs
• Bref historique des réseaux de neurones

• Structure et fonctionnement du perceptron
• Réseaux multicouches
• Entrainer un réseaux de neurones
Machine Learning 2
1
14/12/2021
Non-linear Classification
x2
x1
size
# bedrooms
# floors
age
Machine Learning 3
Car detection
Cars Not a car
Testing:
What is this?
Machine Learning 4
2
14/12/2021
Car detection
pixel 1
Learning
Algorithm
pixel 2
Raw image
pixel 2
Cars pixel 1
“Non”‐Cars
Machine Learning 5
Car detection
pixel 1
Learning
Algorithm
pixel 2
Raw image 50 x 50 pixel images→ 2500 pixels
pixel 2 (7500 if RGB)
pixel 1 intensity
pixel 2 intensity
pixel 2500 intensity
Cars pixel 1 Quadratic features ( ): ≈3 million

“Non”‐Cars features
Machine Learning 6
3
14/12/2021
Neural Networks
• Origins: Algorithms that try to mimic the brain.
• Was very widely used in 80s and early 90s; popularity diminished in late 90s.
• Recent resurgence: State‐of‐the‐art technique for many applications
Machine Learning 7
Neural Networks
• Origin of Neural Networks:
“Artificial neural networks (ANN) or connectionist systems
are computing systems vaguely inspired by the biological
neural networks that constitute animal brains.”
Machine Learning 8
4
14/12/2021
Neural Networks
• Origin of Neural Networks:
“Artificial neural networks (ANN) or connectionist systems
are computing systems vaguely inspired by the biological
neural networks that constitute animal brains.”
Multiple layers in a biological neural network
Machine Learning 9
Perceptron
Machine Learning 10
5
14/12/2021
Perceptron
Machine Learning 11
Perceptron learning rule (weight update)
In this equation:
• wi, j is the connection weight between the ith input neuron and the jth output neuron.
• xi is the ith input value of the current training instance.
• ŷ j is the output of the jth output neuron for the current training instance.
• yj is the target output of the jth output neuron for the current training instance.
• η is the learning rate.
Machine Learning 12
6
14/12/2021
Neuron model: Logistic unit
Machine Learning 13
Neural Networks
Machine Learning 14
7
14/12/2021
Feed-Forward Process
• Input layer units are features
• Working forward through the network, the input
function is applied to compute the input value
• E.g., weighted sum of the input
• The activation function transforms this input
function into a final value
• Typically a nonlinear function (e.g, sigmoid)
Machine Learning 15
Neural Networks
Machine Learning 16
8
14/12/2021
Forward propagation: Vectorized implementation
Machine Learning 17
Neural Networks
• Neural Network learning its own features
Layer 1 Layer 2 Layer 3

Machine Learning 18
9
14/12/2021
Multilayer Perceptron
Layer 1 Layer 2 Layer 3 Layer 4

Machine Learning 19
Non-linear classification
• Example: XOR/XNOR
, are binary (0 or 1).
x2
x2
x1
x1
Machine Learning 20
10
14/12/2021
Machine Learning 21
• Example: OR function
0 0
0 1
1 0
1 1
Machine Learning 22
11
14/12/2021
• Example: Negation function
0
1
Machine Learning 23
• Putting it together: x2
x1
0 0
0 1
1 0
1 1
Machine Learning 24
12
14/12/2021
Multiple output units: One-vs-all
Pedestrian Car Motorcycle Truck
Want , , , etc.
when pedestrian when car when motorcycle
Machine Learning 25
Multiple output units: One-vs-all
Want , , , etc.
when pedestrian when car when motorcycle
Training set:
one of , , ,
pedestrian car motorcycle truck
Machine Learning 26
13
14/12/2021
Cost function
total no. of layers in network
no. of units (not counting bias unit) in
layer
Binary classification Multi‐class classification (K classes)
E.g. , , ,
pedestrian car motorcycle truck
1 output unit K output units
Machine Learning 27
Cost function
• Logistic regression:
• Neural network:
Need code to compute:
‐
‐
Machine Learning 28
14
14/12/2021
Backpropagation algorithm
• Gradient computation
Given one training example ( , ):
Machine Learning 29
Intuition: “error” of node in layer .
For each output unit (layer L = 4)

Machine Learning 30
15
14/12/2021
Set (for all ).
For
Set
Perform forward propagation to compute for
Using , compute
Compute
Machine Learning 31
Forward Propagation
Machine Learning 32
16
14/12/2021
Backpropagation
Machine Learning 33
Network architecture
• Pick a network architecture (connectivity pattern between
neurons)
No. of input units: Dimension of features
No. output units: Number of classes
Reasonable default: 1 hidden layer, or if >1 hidden layer, have same no. of hidden
units in every layer (usually the more the better)
Machine Learning 34
17
14/12/2021
Training a neural network
1. Randomly initialize weights
2. Forward propagation to get for any
3. Compute cost function
4. Backpropagation to compute partial derivatives
5. Use gradient descent or advanced optimization method
with backpropagation to try to minimize as a function
of parameters
Machine Learning 35
Objectifs
 Bref historique des réseaux de neurones

 Structure et fonctionnement du perceptron
 Réseaux multicouches
 Entrainer un réseaux de neurones
Machine Learning 36
18

Chapitre 4

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapitre 4

Uploaded by

Copyright:

Available Formats

14/12/2021

Apprentissage non supervisé

• Bref historique des réseaux de neurones

Cars pixel 1 Quadratic features ( ): ≈3 million

Perceptron learning rule (weight update)

Layer 1 Layer 2 Layer 3

Layer 1 Layer 2 Layer 3 Layer 4

Multiple output units: One-vs-all

Pedestrian Car Motorcycle Truck

Multiple output units: One-vs-all

Layer 1 Layer 2 Layer 3 Layer 4

Training a neural network

 Bref historique des réseaux de neurones

You might also like