management in complexity the exploration of a new paradigm complexity in computing and ai walter...
TRANSCRIPT
Management in complexity
The exploration of a new paradigm
Complexity in computing and AIWalter Baets, PhD, HDR
Associate Dean for Innovation and Social ResponsibilityProfessor Complexity , Knowledge and InnovationEuromed Marseille – Ecole de Management
Chris Langton
Artificial life research
Genetic programming/algorithms
Self-organization (the bee colony)
Interacting (negotiating) agents
Conway’s game of life
One of the earlier artificial life simulations
Simulates behavior of single cells
Rules:
•Any live cell with fewer than two neighbors dies of loneliness•Any live cell with more than three neighbors dies of crowding•Any dead cell with exactly three neighbors comes to life•Any cell with two or three neighbors lives, unchanged to the
next generation
John Holland
Father of genetic programming
Agent-based systems (network)
Individuals have limited characteristics
Individuals optimize their goals
Limited interaction (communication) rules
Complex Adaptive Systems
Artificial Neural Networks
Agent-based systems (network)
Genetic Algorithms
Fuzzy logic
Fuzzy neural networks
ARTIFICIAL NEURAL NETWORKS (ANN) (1)
How does the brain operate?
ARTIFICIAL NEURAL NETWORKS (ANN) (2)
What does an artificial neural network look like?
Out2
Out1
Input Layer Hidden Layer Output Layer
X1
X2
X3
X4
X5
Xn
ARTIFICIAL NEURAL NETWORKS (ANN) (3)
How does an artificial neural network works (gets trained)
NET
TRESHOLDVALUE
X1
X2
X3
X4
Inputs
W1
W2
W3
W4
KNOT
Out-F (net)
Output
ARTIFICIAL NEURAL NETWORKS (ANN) (4)
Comparison to other DSS techniques (advantages)
Able to simulate non-linear behaviour
Has learning behaviour
Non-parametric (no equations)
Fault tolerant (can easily deal with NAs)
Seeking diversity (instead of averages)
Pattern recognition
FUZZY LOGIC (1)
Fuzzy sets and overlapping membership-functions
0 150 185 200
Tall
.7 1
Deg
ree
ofm
embe
rshi
p
Height in cm
FUZZY LOGIC (2)
Representation of the concept size using fuzzy sets
Height in cm
average height
0 150 185 200
Tall
.7
1
Deg
ree
ofm
embe
rshi
p
0.49
short
very tall
FUZZY LOGIC (3)
Fuzzy rules (1)
1
100 90 80 70 60 50 40 30 20 10 0
0
1
45 50 55 60 65 70 75 80 85 90
IF WARMTHEN FAST
STOP
MEDIUM
FAST
0
SLOW
BLAST
COLD COOL
JUST
RIGH
T
WARM HOT
AIR
MO
TOR
SPEE
D
TEMPERATURE IN DEGREES FAHRENHEIT
FUZZY LOGIC (4)
Fuzzy rules (2)
7º 13º 16º 18º 21º 24º 27º 29º 32º10º0
1
0 1
100
90
80
70
60
50
40
30
20
10
IFCOLD,THENSTOP
IF COOL,THENSLOW
IF JUSTRIGHT,THENMEDIUM
IF WARM,THEN FAST
IF HOT,THENBLAST
Fuzzy rulesA
IR M
OT
OR
SP
EE
D
TEMPERATURE IN DEGREES CELSIUS
FUZZY LOGIC (5)
ADVANTAGES:• Smooth behaviour• “Human-like” behaviour• Natural language approach
EXAMPLES:• Sendai Subway• Trading systems• Washing machines, CAM-corders,
micro-waves
FUZZY NEURAL NETWORKS IN MANAGEMENT
Combination of the learning behaviour of neural networks with the fuzziness and the (though fuzzy) rules
Overlapping and vague memberships is a reality in managerial problems
Fuzzy rules is a reality in management
Fuzzy and learning behaviour is very human
Pretty much to be discovered in management sciences
GENETIC ALGORITHMS (1)
GENETIC ALGORITHMS (2)
GENETIC ALGORITHMS (3)
GENETIC ALGORITHMS (4)
GENETIC ALGORITHMS (5)
GENETIC ALGORITHMS (6)
GENETIC ALGORITHMS (7)
GENETIC ALGORITHMS (8)
A beginning of evidenceSome research projects
Complexity and emergent learning in innovation projects: Agents, Sara Lee/DE
Innovation in SME’s: a network structure: ANNs, brainstorm sessions
Telemedecin: a systemic research into the ICT innovations in themedical care market: Agents
Knowledge management at Akzo Nobel: improving the knowledgecreation ability: ANNs, Akzo Nobel
Information ecology: For the moment a conceptual model Agents
Conflict managementAgents
Knowledge management at Bison: contribution to innovationAgents
Complexity in economics
Law of increasing returns (Brian Arthur)
• Characteristics of the information economy (a non-linear dynamic system)
• Phenomenon of increasing returns
• Positive feed-back
• No equilibrium
• Quantum structure of business (WB)
Summary (until now)
• Non - linearity• Dynamic behavior• Dependence on initial conditions• Period doubling• Existence of attractors• Determinism• Emergence at the edge of chaos