Martin Ayvazyan
Activation distribution in
a neural network
20 August 2018
Restricted © 2017 Mentor Graphics Corporation
Contents
 Why are neural networks so important
 Basic network characteristics
 Basic description of the process, formulation of algorithms,
comparison of behavioral characteristics
 Common behavioral characteristics of the process
 A visual representation of the process
 Core characteristics of the software
Activation distribution in a neural network, 20182
Restricted © 2017 Mentor Graphics Corporation
Cell Division and Cancer
Activation distribution in a neural network, 20183
Restricted © 2017 Mentor Graphics Corporation
The global economic crisis prevention /
elimination
Activation distribution in a neural network, 20184
Restricted © 2017 Mentor Graphics Corporation
Traffic control / jams prevention
Activation distribution in a neural network, 20185
Restricted © 2017 Mentor Graphics Corporation
Deep learning
Activation distribution in a neural network, 20186
Restricted © 2017 Mentor Graphics Corporation
Astronomy
Activation distribution in a neural network, 20187
Restricted © 2017 Mentor Graphics Corporation
Basic definitions
ER model
G(N, p)
N – nodes count in the graph
p – component of connectivity
P(G) – the probability that the graph is
connected
Activation distribution in a neural network, 20188
• p =
𝑐∙𝑙𝑛𝑁
𝑁
&& c>1 => graph is connected (c < 1 => graph is disconnected)
• p =
𝑐∙𝑙𝑛𝑁
𝑁
&& (c>3 && N > 100) => 𝑃 𝑁,𝑝 𝐺 ≥ 1 −
1
𝑁

𝑛
𝑁∗ 𝑁−1
2
 𝑃(𝐺0) = 𝑝 𝑛
∙ (1 − 𝑝)
𝑁∙(𝑁−1)
2
−𝑛
 𝑃(𝐺′) =
𝑛
𝑁∙(𝑁−1)
2
∙ 𝑝 𝑛 ∙ (1 − 𝑝)(
𝑁∙ 𝑁−1
2
− 𝑛)
Restricted © 2017 Mentor Graphics Corporation
Small world
Activation distribution in a neural network, 20189
Restricted © 2017 Mentor Graphics Corporation
Clusterization
Activation distribution in a neural network, 201810
𝐶𝑗 𝑞 𝑗 =
𝑡𝑗
𝑞 𝑗 ∙ (𝑞 𝑗 − 1)/2
𝐶 =
1
𝑁
𝑗
𝐶𝑗
Restricted © 2017 Mentor Graphics Corporation
Degree distribution
Activation distribution in a neural network, 201811
Restricted © 2017 Mentor Graphics Corporation
Algorithm A
 Input: a network, list of active nodes in the network, μ, λ
 Step 1: If there is no an active node in the network – exit
 Step 2: Randomly is selected a node
 Step 3: If the selected node doesn’t active – go to Step 2
 Step 4: Randomly is selected an adjacent node and in case if the
neighbor doesn’t active – activate it with
𝛌
(𝛌+ 𝛍)
probability
 Step 5: The temporary selected node is deactivated with
𝜇
(𝜆+ 𝜇)
probability
 Step 6: Go to Step 1
Activation distribution in a neural network, 201812
Restricted © 2017 Mentor Graphics Corporation
Algorithm A
𝑁=1024, 𝑝=0.5, μ=0.0
Activation distribution in a neural network, 201813
Restricted © 2017 Mentor Graphics Corporation
Algorithm A
Activation distribution in a neural network, 201814
P={0.5, 0.1, 0.05} λ = 0.005 P={0.5, 0.1, 0.05} λ = 0.05
Restricted © 2017 Mentor Graphics Corporation
Algorithm B
 Input: a network, list of active nodes in the network, μ, λ
 Step 1: If there is no an active node in the network – exit
 Step 2: Randomly is selected a node
 Step 3: If the selected node doesn’t active – go to Step 2
 Step 4: Each adjacent inactive node is activated with
𝛌
(𝛌+ 𝛍)
probability
 Step 5: The temporary selected node is deactivated with
𝜇
(𝜆+ 𝜇)
probability
 Step 6: Go to Step 1
Activation distribution in a neural network, 201815
Restricted © 2017 Mentor Graphics Corporation
Algorithm B
Activation distribution in a neural network, 201816
N=1024, p=0.5, μ=0.0
Restricted © 2017 Mentor Graphics Corporation
Algorithm B
Activation distribution in a neural network, 201817
N=1024, P = {0.5, 0.1, 0.05}, λ = 0.05 N=1024, P = {0.5, 0.1, 0.05}, λ = 0.0005
Restricted © 2017 Mentor Graphics Corporation
Algorithm C
 Input: a network, list of active nodes in the network, μ, λ
 Step 1: If there is no an active node in the network – exit
 Step 2: For each active note of the current stage:
— Step 2.1: Randomly is selected an adjacent node and in case if the
neighbor doesn’t active – activate it with
𝛌
(𝛌+ 𝛍)
probability
— Step 2.2: The node is deactivated with
𝜇
(𝜆+ 𝜇)
probability
 Step 3: Go to Step 1
Activation distribution in a neural network, 201818
Restricted © 2017 Mentor Graphics Corporation
Algorithm C
Activation distribution in a neural network, 201819
N=1024, p=0.5, μ=0.0
Restricted © 2017 Mentor Graphics Corporation
Algorithm C
Activation distribution in a neural network, 201820
N=1024, P = {0.5, 0.1, 0.05}, λ = 0.0005 N=1024, P = {0.5, 0.1, 0.05}, λ = 0.0005
Restricted © 2017 Mentor Graphics Corporation
Common behavioral characteristics of the process
ER(n,p), active_nodes_count = 1
 The probability that in the current step will be activated at least one
node:
Activation distribution in a neural network, 201821
‘A’ ‘B’ ‘C’
1
𝑛
∙ 𝜆
1
𝑛
∙ (1 − 1 − 𝜆 𝑝∙ 𝑛−1 )
𝜆
Restricted © 2017 Mentor Graphics Corporation
Common behavioral characteristics of the process
 The probability that in the current step will be activated m / s.t {1 < m <= p ∙ (n − 1)}
nodes:
 The probability that after the current step in the network there will be no active nodes:
Activation distribution in a neural network, 201822
‘A’ ‘B’ ‘C’
0 1
𝑛
∙ (
𝑚
𝑝 ∙ 𝑛 − 1
∙ 𝜆 𝑚
∙ 1 − 𝜆 𝑝∙ 𝑛−1 −𝑚
) 0
‘A’ ‘B’ ‘C’
μ∙(
1
𝑛
∙ (1 − 𝜆)) μ∙(
1
𝑛
∙ 1 − 𝜆 𝑝∙(𝑛−1)) μ ∙ 1 − 𝜆
Restricted © 2017 Mentor Graphics Corporation
A visual representation of the process
ER(n, p) / n=65, p = 005, active_nodes_count=65, 𝜆 = 0.314159265, μ=1.0
Activation distribution in a neural network, 201823
0 20 30
50 500 1500
Restricted © 2017 Mentor Graphics Corporation
Core characteristics of the software
 Accuracy
 Efficiency
 Platform independency
Activation distribution in a neural network, 201824
Restricted © 2017 Mentor Graphics Corporation
www.mentor.com

Activation distribution in a neural network

  • 1.
    Martin Ayvazyan Activation distributionin a neural network 20 August 2018
  • 2.
    Restricted © 2017Mentor Graphics Corporation Contents  Why are neural networks so important  Basic network characteristics  Basic description of the process, formulation of algorithms, comparison of behavioral characteristics  Common behavioral characteristics of the process  A visual representation of the process  Core characteristics of the software Activation distribution in a neural network, 20182
  • 3.
    Restricted © 2017Mentor Graphics Corporation Cell Division and Cancer Activation distribution in a neural network, 20183
  • 4.
    Restricted © 2017Mentor Graphics Corporation The global economic crisis prevention / elimination Activation distribution in a neural network, 20184
  • 5.
    Restricted © 2017Mentor Graphics Corporation Traffic control / jams prevention Activation distribution in a neural network, 20185
  • 6.
    Restricted © 2017Mentor Graphics Corporation Deep learning Activation distribution in a neural network, 20186
  • 7.
    Restricted © 2017Mentor Graphics Corporation Astronomy Activation distribution in a neural network, 20187
  • 8.
    Restricted © 2017Mentor Graphics Corporation Basic definitions ER model G(N, p) N – nodes count in the graph p – component of connectivity P(G) – the probability that the graph is connected Activation distribution in a neural network, 20188 • p = 𝑐∙𝑙𝑛𝑁 𝑁 && c>1 => graph is connected (c < 1 => graph is disconnected) • p = 𝑐∙𝑙𝑛𝑁 𝑁 && (c>3 && N > 100) => 𝑃 𝑁,𝑝 𝐺 ≥ 1 − 1 𝑁  𝑛 𝑁∗ 𝑁−1 2  𝑃(𝐺0) = 𝑝 𝑛 ∙ (1 − 𝑝) 𝑁∙(𝑁−1) 2 −𝑛  𝑃(𝐺′) = 𝑛 𝑁∙(𝑁−1) 2 ∙ 𝑝 𝑛 ∙ (1 − 𝑝)( 𝑁∙ 𝑁−1 2 − 𝑛)
  • 9.
    Restricted © 2017Mentor Graphics Corporation Small world Activation distribution in a neural network, 20189
  • 10.
    Restricted © 2017Mentor Graphics Corporation Clusterization Activation distribution in a neural network, 201810 𝐶𝑗 𝑞 𝑗 = 𝑡𝑗 𝑞 𝑗 ∙ (𝑞 𝑗 − 1)/2 𝐶 = 1 𝑁 𝑗 𝐶𝑗
  • 11.
    Restricted © 2017Mentor Graphics Corporation Degree distribution Activation distribution in a neural network, 201811
  • 12.
    Restricted © 2017Mentor Graphics Corporation Algorithm A  Input: a network, list of active nodes in the network, μ, λ  Step 1: If there is no an active node in the network – exit  Step 2: Randomly is selected a node  Step 3: If the selected node doesn’t active – go to Step 2  Step 4: Randomly is selected an adjacent node and in case if the neighbor doesn’t active – activate it with 𝛌 (𝛌+ 𝛍) probability  Step 5: The temporary selected node is deactivated with 𝜇 (𝜆+ 𝜇) probability  Step 6: Go to Step 1 Activation distribution in a neural network, 201812
  • 13.
    Restricted © 2017Mentor Graphics Corporation Algorithm A 𝑁=1024, 𝑝=0.5, μ=0.0 Activation distribution in a neural network, 201813
  • 14.
    Restricted © 2017Mentor Graphics Corporation Algorithm A Activation distribution in a neural network, 201814 P={0.5, 0.1, 0.05} λ = 0.005 P={0.5, 0.1, 0.05} λ = 0.05
  • 15.
    Restricted © 2017Mentor Graphics Corporation Algorithm B  Input: a network, list of active nodes in the network, μ, λ  Step 1: If there is no an active node in the network – exit  Step 2: Randomly is selected a node  Step 3: If the selected node doesn’t active – go to Step 2  Step 4: Each adjacent inactive node is activated with 𝛌 (𝛌+ 𝛍) probability  Step 5: The temporary selected node is deactivated with 𝜇 (𝜆+ 𝜇) probability  Step 6: Go to Step 1 Activation distribution in a neural network, 201815
  • 16.
    Restricted © 2017Mentor Graphics Corporation Algorithm B Activation distribution in a neural network, 201816 N=1024, p=0.5, μ=0.0
  • 17.
    Restricted © 2017Mentor Graphics Corporation Algorithm B Activation distribution in a neural network, 201817 N=1024, P = {0.5, 0.1, 0.05}, λ = 0.05 N=1024, P = {0.5, 0.1, 0.05}, λ = 0.0005
  • 18.
    Restricted © 2017Mentor Graphics Corporation Algorithm C  Input: a network, list of active nodes in the network, μ, λ  Step 1: If there is no an active node in the network – exit  Step 2: For each active note of the current stage: — Step 2.1: Randomly is selected an adjacent node and in case if the neighbor doesn’t active – activate it with 𝛌 (𝛌+ 𝛍) probability — Step 2.2: The node is deactivated with 𝜇 (𝜆+ 𝜇) probability  Step 3: Go to Step 1 Activation distribution in a neural network, 201818
  • 19.
    Restricted © 2017Mentor Graphics Corporation Algorithm C Activation distribution in a neural network, 201819 N=1024, p=0.5, μ=0.0
  • 20.
    Restricted © 2017Mentor Graphics Corporation Algorithm C Activation distribution in a neural network, 201820 N=1024, P = {0.5, 0.1, 0.05}, λ = 0.0005 N=1024, P = {0.5, 0.1, 0.05}, λ = 0.0005
  • 21.
    Restricted © 2017Mentor Graphics Corporation Common behavioral characteristics of the process ER(n,p), active_nodes_count = 1  The probability that in the current step will be activated at least one node: Activation distribution in a neural network, 201821 ‘A’ ‘B’ ‘C’ 1 𝑛 ∙ 𝜆 1 𝑛 ∙ (1 − 1 − 𝜆 𝑝∙ 𝑛−1 ) 𝜆
  • 22.
    Restricted © 2017Mentor Graphics Corporation Common behavioral characteristics of the process  The probability that in the current step will be activated m / s.t {1 < m <= p ∙ (n − 1)} nodes:  The probability that after the current step in the network there will be no active nodes: Activation distribution in a neural network, 201822 ‘A’ ‘B’ ‘C’ 0 1 𝑛 ∙ ( 𝑚 𝑝 ∙ 𝑛 − 1 ∙ 𝜆 𝑚 ∙ 1 − 𝜆 𝑝∙ 𝑛−1 −𝑚 ) 0 ‘A’ ‘B’ ‘C’ μ∙( 1 𝑛 ∙ (1 − 𝜆)) μ∙( 1 𝑛 ∙ 1 − 𝜆 𝑝∙(𝑛−1)) μ ∙ 1 − 𝜆
  • 23.
    Restricted © 2017Mentor Graphics Corporation A visual representation of the process ER(n, p) / n=65, p = 005, active_nodes_count=65, 𝜆 = 0.314159265, μ=1.0 Activation distribution in a neural network, 201823 0 20 30 50 500 1500
  • 24.
    Restricted © 2017Mentor Graphics Corporation Core characteristics of the software  Accuracy  Efficiency  Platform independency Activation distribution in a neural network, 201824
  • 25.
    Restricted © 2017Mentor Graphics Corporation www.mentor.com