Decision%20Support%20System:

1
Chapter 8 Models and Decision Support

• Decision Support System
• DSS is an interactive IT system that creates new
  information on demand to help you make
  nonstructural decisions.

2

• Components of Decision Support System (p.295)

3

• Why do we use models? (p.290)
• Understand process
• Optimization
• Prediction
• Simulation
• Problems when using models
• Complexity of model
• Cost of model
• Error of model

4

• Popular Decision Support Models
• Prediction regression model, other statistical
  model.
• What-If analysis simulation
• Optimization models Optimization models seek the
  best solution where decision need to be made in a
  constrained or limited resource environment. An
  optimization model has its objective function,
  decision variables and constraints.
• Heuristic methods A heuristic is a practical
  and quick method based on intuitive and plausible
  arguments that likely to (but not guaranteed to)
  lead to a solution that is approximately optimal
  or near optimal.
• K. G. Murty, 1995, Operations Research.

5

• Why many managers still seem to prefer heuristic
  method?
• There are no optimization algorithms to find the
  optimum solutions to many large scale problems in
  real business world within acceptable computer
  time.
• Because of errors in the data, an optimal
  solution can only be a good solution for the real
  problem and serves the same purpose as a good but
  simple heuristic.

6

• Example Optimum solution and a good solution by
  heuristic. Capacity is 35 hours.
• Optimum solution projects 2, 5, and 7. Profit
  351.
• Greedy heuristic projects 7, 5, and 1. Profit
  348.

Project Hours Profit Profit/Hour
1 3 21 7
2 4 24 6
3 3 12 4
4 21 168 8
5 15 135 9
6 13 26 1
7 16 192 12
8 20 200 10
9 40 800 Eliminate
7

• Data Mining
• A collection of data analysis techniques for
  extremely large dataset.
• Properly applied, data mining can reveal hidden
  relationships and information buried within the
  data warehouse..
• G. M. Marakas, 2003, Modern Data Warehousing,
  Mining, and Visualization.

8

• The purpose to use EIS
• Display transaction data in a form of overall
  picture.
• Provide easy access to corporate data for
  executives.
• Advantages of EIS
• Top executive can retrieve data without waiting
  for report.
• Data of all functional areas are retrieved from
  the same system a better big picture.
• Drill down to get more detailed data.
• Comparison of time series and cross-sectional
  data.
• The data is nonintrusive (a copy of the
  transaction data).