اسلاید روز چهارم از کارگاه ۷ روزه دادههای بزرگ و یادگیری ماشین که شامل مقدمه ای بر شبکههای عصبی مصنوعی و یک نمونه پیاده سازی ساده به زبان جاوا است. این دوره به همت ایسیام دانشگاه تهران برگزار میشود
زمان هر جلسه ۲ ساعت است
61. 61
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻧﻮﺭرﻭوﻥن
1836 - Discovery of the neural cell
of the brain, the neuron
62. 62
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻧﻮﺭرﻭوﻥن
1836 - Discovery of the neural cell
of the brain, the neuron
63. 63
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻧﻮﺭرﻭوﻥن
1836 - Discovery of the neural cell
of the brain, the neuron
1897 - Synapse concept introduced
64. 64
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻧﻮﺭرﻭوﻥن
1836 - Discovery of the neural cell
of the brain, the neuron
1897 - Synapse concept introduced
1943 First mathematical
representation of neuron
65. 65
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
1836 - Discovery of the neural cell
of the brain, the neuron
1897 - Synapse concept introduced
1943 First mathematical
representation of neuron
66. 66
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
The neuron calculates a
weighted sum of inputs
and compares it to a threshold.
If the sum is higher than the
threshold, the output is set to 1,
otherwise to -1.
67. 67
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
(Perceptron) ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
Artificial neuron models are at their core simplified models based on biological
neurons.
68. 68
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
(Perceptron) ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
Artificial neuron models are at their core simplified models based on biological
neurons.
Many Inputs
Individual Weighted
69. 69
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
(Perceptron) ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
Artificial neuron models are at their core simplified models based on biological
neurons.
Many Inputs
Individual Weighted
Amplify/Deamplify
70. 70
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
(Perceptron) ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
Artificial neuron models are at their core simplified models based on biological
neurons.
Many Inputs
Individual Weighted
Amplify/Deamplify
Adds the weighted
Signals together
71. 71
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
(Perceptron) ﻣﺼﻨﻮﻋﯽ ﻧﻮﺭرﻭوﻥن
Artificial neuron models are at their core simplified models based on biological
neurons.
Many Inputs
Individual Weighted
Amplify/Deamplify
Adds the weighted
Signals together
Converts the input into a more useful output
83. 83
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ
The input units, are designed to receive
various forms of information from the outside
world that the network will attempt to learn
about, recognize, or otherwise process.
84. 84
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ
The higher the weight, the more influence
one unit has on another.
The input units, are designed to receive
various forms of information from the outside
world that the network will attempt to learn
about, recognize, or otherwise process.
85. 85
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ
The higher the weight, the more influence
one unit has on another.
The input units, are designed to receive
various forms of information from the outside
world that the network will attempt to learn
about, recognize, or otherwise process.
Most neural networks are fully connected.
86. 86
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ
The higher the weight, the more influence
one unit has on another.
The input units, are designed to receive
various forms of information from the outside
world that the network will attempt to learn
about, recognize, or otherwise process.
Most neural networks are fully connected.
In between the input units and output units
are one or more layers of hidden units, which,
together, form the majority of the artificial
brain.
87. 87
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ
The higher the weight, the more influence
one unit has on another.
The input units, are designed to receive
various forms of information from the outside
world that the network will attempt to learn
about, recognize, or otherwise process.
Most neural networks are fully connected.
In between the input units and output units
are one or more layers of hidden units, which,
together, form the majority of the artificial
brain.
The strength (weight) of the connection
between any two units is gradually adjusted
as the network learns.
89. 89
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
For a neural network to learn, there has to
be an element of feedback involved—just
as children learn by being told what they're
doing right or wrong.
92. 92
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
…ﺍاﺳﺖ ﮐﺮﺩدﻥن ﭘﺮ ﺍاﺯز ﮐﺮﺩدﻥن ﻧﯿﮑﻮ ﮐﺎﺭر
In fact, we all use feedback, all the time.
93. 93
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
In fact, we all use feedback, all the time.
Neural networks learn things in
exactly the same way, typically by a
feedback process called
backpropagation
94. 94
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
In fact, we all use feedback, all the time.
Neural networks learn things in
exactly the same way, typically by a
feedback process called
backpropagation
95. 95
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
It is believed that during the learning
process the brain's neural structure is
altered, increasing or decreasing the
strength of it's synaptic connections
depending on their activity.
96. 96
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
It is believed that during the learning
process the brain's neural structure is
altered, increasing or decreasing the
strength of it's synaptic connections
depending on their activity.
More relevant information will have
stronger synaptic connections and
less relevant information will gradually
have it's synaptic connections
weaken, making it harder to recall.
97. 97
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﻣﺼﻨﻮﻋﯽ ﻋﺼﺒﯽ ﺷﺒﮑﻪ ﺩدﺭر ﯾﺎﺩدﮔﯿﺮﯼی
It is believed that during the learning
process the brain's neural structure is
altered, increasing or decreasing the
strength of it's synaptic connections
depending on their activity.
More relevant information will have
stronger synaptic connections and
less relevant information will gradually
have it's synaptic connections
weaken, making it harder to recall.
117. 117
ﺍاﻥنﺮﺗﻬ ﺩدﺍاﻧﺸﮕﺎﻩه ACM - ۱۳۹۵ ﺗﺎﺑﺴﺘﺎﻥن - ﻣﺎﺷﯿﻦﯼیﯾﺎﺩدﮔﯿﺮ ﻭو ﺑﺰﺭرﮒگ ﻫﺎﯼیﻩهﺩدﺍاﺩد ﻫﺎﯼیﺩدﮐﺎﺭرﺑﺮ ﺑﺮ ﮔﺬﺭرﯼی
ﺍاﺟﺮﺍا
mvn compile
mvn exec:java -
Dexec.mainClass=“ir.ac.ut.acm.PerceptronLearning.PerceptronLearningRule”
Run the PerceptronLearningRule class by using the following commands: