Support System Management: Ing. Ridho Rahmadi, M.SC

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 29

Support System Management

Ing. Ridho Rahmadi, M.Sc

Magister Teknik Informatika


Universitas Islam Indonesia
ridho.rahmadi@uii.ac.id

October 13, 2017

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 1 / 29


Acknowledgment

Thanks to Dr. Sri Kusumadewi!


Most contents in these slides are based on her material.

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 2 / 29


Contact

ridho.rahmadi@uii.ac.id
08112951304

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 3 / 29


Recap 1

Why we need a model?


Making a decision is not an easy task

Four phases of decision making (Turban, 2005)


1 Problem identification → a formal problem statement
2 Model design
3 Model selection
4 Implementation

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 4 / 29


Modeling

A modelling is an attempt to
analyze a real-world problem by Why
simulating it. 1 Much easier
2 Cost and time saving
3 Risk minimizing
4 A proper approach to
uncertainty
5 Explore (enormous) possible
solutions

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 5 / 29


Important discussions in DSS

Problem identification
Variables identification
Forecasting
Multiple decision models
Model category selection
Model management
Knowledge-based model

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 6 / 29


Problem identification

Collecting much information


Understanding organization
culture and pattern
Surrogates: business process
& problem causes analyses

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 7 / 29


Business process analysis

Business process is a set of


activities designed to accomplish
an organizational goal
Tools to represent
Flowchart
Block diagram
Activity diagram

Motivation: To understand the


organization

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 8 / 29


Business process illustration

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 9 / 29


Problem causes

To figure out possible causes of


the problem
Tool
Fishbone diagram (also
called Ishikawa or
cause-effect diagram)

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 10 / 29


Fishbone example

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 11 / 29


Variables identification

To identify relevant variables


Tools
Bayesian network
(probabilistic)
Influence diagram

Motivation: to observe relations


among variables

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 12 / 29


Bayesian Network

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 13 / 29


Bayesian Network cont’d

Joint probability function


P(G , S, R) = P(G | S, R) P(S | R) P(R)

Example: what is the probability that it is raining, given the grass is wet?

P
P(G = T , R = T ) S∈{T ,F } P(G = T , S, R = T )
P(R = T | G = T ) = = P
P(G = T ) S,R∈{T ,F } P(G = T , S, R)

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 14 / 29


Bayesian Network cont’d

Example: what is the probability that it is raining, given the grass is wet?
Answer:
P
P(G = T , R = T ) S∈{T ,F } P(G = T , S, R = T )
P(R = T | G = T ) = = P
P(G = T ) S,R∈{T ,F } P(G = T , S, R)

where, for example:

P(G = T , S = T , R = T ) = P(G = T | S = T , R = T ) P(S = T | R = T ) P(R = T )


= 0.99 × 0.01 × 0.2
= 0.00198

P(G = T , S = F , R = T ) = P(G = T | S = F , R = T ) P(S = F | R = T ) P(R = T )


= 0.8 × 0.99 × 0.2
= 0.1584

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 15 / 29


0.00198TTT + 0.1584TFT
P(R = T | G = T ) =
0.00198TTT + 0.288TTF + 0.1584TFT + 0.0TFF
≈ 0.35

So, the probability of raining, given that the grass is wet, is approximately
0.35.

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 16 / 29


Influence diagram

A generalization of Bayesian
Network; not only probabilistic,
but also decision making.
Node types
Decision (rectangle)
Uncertainty/chance (oval)
General/deterministic
(rounded rectangle)
Objective/value (diamond)

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 17 / 29


Influence diagram cont’d

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 18 / 29


Influence diagram cont’d

An example:
Suppose a company director would like to make a decision based on the
following criteria:
price
investation
support to company productivity
product availability

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 19 / 29


Influence diagram cont’d

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 20 / 29


Model management

Practically,
A DSS can employ a multiple models
Each model performs different tasks in decision making
Thus, the models need to be managed → model base management system.

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 21 / 29


DSS components

Turban, et al (2005):
Data management
Model management
External models
Knowledge-based subsystem
User interface

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 22 / 29


Model Management

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 23 / 29


Model Management

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 24 / 29


Data management system

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 25 / 29


Model management system

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 26 / 29


User interface system

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 27 / 29


Knowledge-based system

knowledge-based subsystem
to support other subsytems
Some artificial intelligence
approaches are suitable
E.g., Expert system →
qualitative inference

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 28 / 29


Any questions?

Ing. Ridho Rahmadi, M.Sc SPK-BI October 13, 2017 29 / 29

You might also like